diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
deleted file mode 100644--- "a/last-checkpoint/trainer_state.json"
+++ /dev/null
@@ -1,11812 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 14.0,
-  "global_step": 972622,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.01,
-      "learning_rate": 5.140331037318804e-06,
-      "loss": 9.6819,
-      "step": 500
-    },
-    {
-      "epoch": 0.01,
-      "learning_rate": 1.0280662074637608e-05,
-      "loss": 8.0932,
-      "step": 1000
-    },
-    {
-      "epoch": 0.02,
-      "learning_rate": 1.542099311195641e-05,
-      "loss": 6.9551,
-      "step": 1500
-    },
-    {
-      "epoch": 0.03,
-      "learning_rate": 2.0561324149275216e-05,
-      "loss": 6.6147,
-      "step": 2000
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 2.570165518659402e-05,
-      "loss": 6.3513,
-      "step": 2500
-    },
-    {
-      "epoch": 0.04,
-      "learning_rate": 3.084198622391282e-05,
-      "loss": 6.0637,
-      "step": 3000
-    },
-    {
-      "epoch": 0.05,
-      "learning_rate": 3.5982317261231625e-05,
-      "loss": 5.6379,
-      "step": 3500
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 4.112264829855043e-05,
-      "loss": 5.1989,
-      "step": 4000
-    },
-    {
-      "epoch": 0.06,
-      "learning_rate": 4.6262979335869235e-05,
-      "loss": 4.867,
-      "step": 4500
-    },
-    {
-      "epoch": 0.07,
-      "learning_rate": 5.140331037318804e-05,
-      "loss": 4.6153,
-      "step": 5000
-    },
-    {
-      "epoch": 0.08,
-      "learning_rate": 5.6543641410506844e-05,
-      "loss": 4.4195,
-      "step": 5500
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.168397244782565e-05,
-      "loss": 4.2546,
-      "step": 6000
-    },
-    {
-      "epoch": 0.09,
-      "learning_rate": 6.682430348514444e-05,
-      "loss": 4.1156,
-      "step": 6500
-    },
-    {
-      "epoch": 0.1,
-      "learning_rate": 7.196463452246325e-05,
-      "loss": 3.9908,
-      "step": 7000
-    },
-    {
-      "epoch": 0.11,
-      "learning_rate": 7.710496555978204e-05,
-      "loss": 3.8759,
-      "step": 7500
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 8.224529659710087e-05,
-      "loss": 3.7715,
-      "step": 8000
-    },
-    {
-      "epoch": 0.12,
-      "learning_rate": 8.738562763441966e-05,
-      "loss": 3.6865,
-      "step": 8500
-    },
-    {
-      "epoch": 0.13,
-      "learning_rate": 9.252595867173847e-05,
-      "loss": 3.6037,
-      "step": 9000
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 9.766628970905726e-05,
-      "loss": 3.5251,
-      "step": 9500
-    },
-    {
-      "epoch": 0.14,
-      "learning_rate": 9.997164799900302e-05,
-      "loss": 3.4598,
-      "step": 10000
-    },
-    {
-      "epoch": 0.15,
-      "learning_rate": 9.991982511073378e-05,
-      "loss": 3.394,
-      "step": 10500
-    },
-    {
-      "epoch": 0.16,
-      "learning_rate": 9.986789836898105e-05,
-      "loss": 3.3329,
-      "step": 11000
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 9.98159716272283e-05,
-      "loss": 3.2819,
-      "step": 11500
-    },
-    {
-      "epoch": 0.17,
-      "learning_rate": 9.976404488547557e-05,
-      "loss": 3.2306,
-      "step": 12000
-    },
-    {
-      "epoch": 0.18,
-      "learning_rate": 9.971222199720634e-05,
-      "loss": 3.1876,
-      "step": 12500
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 9.966029525545361e-05,
-      "loss": 3.1435,
-      "step": 13000
-    },
-    {
-      "epoch": 0.19,
-      "learning_rate": 9.960836851370088e-05,
-      "loss": 3.1032,
-      "step": 13500
-    },
-    {
-      "epoch": 0.2,
-      "learning_rate": 9.955644177194815e-05,
-      "loss": 3.07,
-      "step": 14000
-    },
-    {
-      "epoch": 0.21,
-      "learning_rate": 9.950461888367891e-05,
-      "loss": 3.0312,
-      "step": 14500
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 9.945269214192618e-05,
-      "loss": 2.9979,
-      "step": 15000
-    },
-    {
-      "epoch": 0.22,
-      "learning_rate": 9.940076540017345e-05,
-      "loss": 2.9722,
-      "step": 15500
-    },
-    {
-      "epoch": 0.23,
-      "learning_rate": 9.93488386584207e-05,
-      "loss": 2.945,
-      "step": 16000
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 9.929701577015148e-05,
-      "loss": 2.9172,
-      "step": 16500
-    },
-    {
-      "epoch": 0.24,
-      "learning_rate": 9.924519288188225e-05,
-      "loss": 2.8928,
-      "step": 17000
-    },
-    {
-      "epoch": 0.25,
-      "learning_rate": 9.919326614012952e-05,
-      "loss": 2.8699,
-      "step": 17500
-    },
-    {
-      "epoch": 0.26,
-      "learning_rate": 9.914133939837677e-05,
-      "loss": 2.8491,
-      "step": 18000
-    },
-    {
-      "epoch": 0.27,
-      "learning_rate": 9.908941265662404e-05,
-      "loss": 2.8275,
-      "step": 18500
-    },
-    {
-      "epoch": 0.27,
-      "learning_rate": 9.903748591487131e-05,
-      "loss": 2.8038,
-      "step": 19000
-    },
-    {
-      "epoch": 0.28,
-      "learning_rate": 9.898555917311856e-05,
-      "loss": 2.787,
-      "step": 19500
-    },
-    {
-      "epoch": 0.29,
-      "learning_rate": 9.893363243136583e-05,
-      "loss": 2.768,
-      "step": 20000
-    },
-    {
-      "epoch": 0.3,
-      "learning_rate": 9.88818095430966e-05,
-      "loss": 2.7538,
-      "step": 20500
-    },
-    {
-      "epoch": 0.3,
-      "learning_rate": 9.882988280134387e-05,
-      "loss": 2.7372,
-      "step": 21000
-    },
-    {
-      "epoch": 0.31,
-      "learning_rate": 9.877795605959113e-05,
-      "loss": 2.7222,
-      "step": 21500
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 9.87260293178384e-05,
-      "loss": 2.7134,
-      "step": 22000
-    },
-    {
-      "epoch": 0.32,
-      "learning_rate": 9.867410257608566e-05,
-      "loss": 2.6956,
-      "step": 22500
-    },
-    {
-      "epoch": 0.33,
-      "learning_rate": 9.862227968781644e-05,
-      "loss": 2.6807,
-      "step": 23000
-    },
-    {
-      "epoch": 0.34,
-      "learning_rate": 9.857035294606369e-05,
-      "loss": 2.671,
-      "step": 23500
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 9.851842620431096e-05,
-      "loss": 2.6575,
-      "step": 24000
-    },
-    {
-      "epoch": 0.35,
-      "learning_rate": 9.846649946255823e-05,
-      "loss": 2.6476,
-      "step": 24500
-    },
-    {
-      "epoch": 0.36,
-      "learning_rate": 9.8414676574289e-05,
-      "loss": 2.6363,
-      "step": 25000
-    },
-    {
-      "epoch": 0.37,
-      "learning_rate": 9.836274983253626e-05,
-      "loss": 2.6247,
-      "step": 25500
-    },
-    {
-      "epoch": 0.37,
-      "learning_rate": 9.831082309078353e-05,
-      "loss": 2.6118,
-      "step": 26000
-    },
-    {
-      "epoch": 0.38,
-      "learning_rate": 9.82588963490308e-05,
-      "loss": 2.6042,
-      "step": 26500
-    },
-    {
-      "epoch": 0.39,
-      "learning_rate": 9.820707346076157e-05,
-      "loss": 2.5936,
-      "step": 27000
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 9.815514671900883e-05,
-      "loss": 2.5828,
-      "step": 27500
-    },
-    {
-      "epoch": 0.4,
-      "learning_rate": 9.810321997725609e-05,
-      "loss": 2.5737,
-      "step": 28000
-    },
-    {
-      "epoch": 0.41,
-      "learning_rate": 9.805129323550336e-05,
-      "loss": 2.5671,
-      "step": 28500
-    },
-    {
-      "epoch": 0.42,
-      "learning_rate": 9.799947034723412e-05,
-      "loss": 2.5575,
-      "step": 29000
-    },
-    {
-      "epoch": 0.42,
-      "learning_rate": 9.794754360548139e-05,
-      "loss": 2.5498,
-      "step": 29500
-    },
-    {
-      "epoch": 0.43,
-      "learning_rate": 9.789561686372866e-05,
-      "loss": 2.5449,
-      "step": 30000
-    },
-    {
-      "epoch": 0.44,
-      "learning_rate": 9.784369012197591e-05,
-      "loss": 2.5361,
-      "step": 30500
-    },
-    {
-      "epoch": 0.45,
-      "learning_rate": 9.77918672337067e-05,
-      "loss": 2.5307,
-      "step": 31000
-    },
-    {
-      "epoch": 0.45,
-      "learning_rate": 9.774004434543746e-05,
-      "loss": 2.523,
-      "step": 31500
-    },
-    {
-      "epoch": 0.46,
-      "learning_rate": 9.768811760368473e-05,
-      "loss": 2.5129,
-      "step": 32000
-    },
-    {
-      "epoch": 0.47,
-      "learning_rate": 9.763619086193198e-05,
-      "loss": 2.509,
-      "step": 32500
-    },
-    {
-      "epoch": 0.48,
-      "learning_rate": 9.758426412017925e-05,
-      "loss": 2.4989,
-      "step": 33000
-    },
-    {
-      "epoch": 0.48,
-      "learning_rate": 9.753233737842652e-05,
-      "loss": 2.4968,
-      "step": 33500
-    },
-    {
-      "epoch": 0.49,
-      "learning_rate": 9.748051449015729e-05,
-      "loss": 2.4901,
-      "step": 34000
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 9.742858774840455e-05,
-      "loss": 2.4799,
-      "step": 34500
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 9.737666100665181e-05,
-      "loss": 2.4764,
-      "step": 35000
-    },
-    {
-      "epoch": 0.51,
-      "learning_rate": 9.732473426489909e-05,
-      "loss": 2.4707,
-      "step": 35500
-    },
-    {
-      "epoch": 0.52,
-      "learning_rate": 9.727280752314635e-05,
-      "loss": 2.4663,
-      "step": 36000
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 9.722098463487713e-05,
-      "loss": 2.4592,
-      "step": 36500
-    },
-    {
-      "epoch": 0.53,
-      "learning_rate": 9.716905789312438e-05,
-      "loss": 2.4528,
-      "step": 37000
-    },
-    {
-      "epoch": 0.54,
-      "learning_rate": 9.711713115137165e-05,
-      "loss": 2.4517,
-      "step": 37500
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 9.706520440961892e-05,
-      "loss": 2.447,
-      "step": 38000
-    },
-    {
-      "epoch": 0.55,
-      "learning_rate": 9.701338152134968e-05,
-      "loss": 2.4342,
-      "step": 38500
-    },
-    {
-      "epoch": 0.56,
-      "learning_rate": 9.696145477959695e-05,
-      "loss": 2.4355,
-      "step": 39000
-    },
-    {
-      "epoch": 0.57,
-      "learning_rate": 9.69095280378442e-05,
-      "loss": 2.4292,
-      "step": 39500
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 9.685760129609149e-05,
-      "loss": 2.426,
-      "step": 40000
-    },
-    {
-      "epoch": 0.58,
-      "learning_rate": 9.680567455433874e-05,
-      "loss": 2.4177,
-      "step": 40500
-    },
-    {
-      "epoch": 0.59,
-      "learning_rate": 9.6753747812586e-05,
-      "loss": 2.4136,
-      "step": 41000
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 9.670192492431678e-05,
-      "loss": 2.4099,
-      "step": 41500
-    },
-    {
-      "epoch": 0.6,
-      "learning_rate": 9.664999818256403e-05,
-      "loss": 2.4072,
-      "step": 42000
-    },
-    {
-      "epoch": 0.61,
-      "learning_rate": 9.659807144081131e-05,
-      "loss": 2.402,
-      "step": 42500
-    },
-    {
-      "epoch": 0.62,
-      "learning_rate": 9.654614469905857e-05,
-      "loss": 2.3975,
-      "step": 43000
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 9.649432181078935e-05,
-      "loss": 2.3914,
-      "step": 43500
-    },
-    {
-      "epoch": 0.63,
-      "learning_rate": 9.64423950690366e-05,
-      "loss": 2.3901,
-      "step": 44000
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 9.639046832728387e-05,
-      "loss": 2.388,
-      "step": 44500
-    },
-    {
-      "epoch": 0.65,
-      "learning_rate": 9.633854158553114e-05,
-      "loss": 2.3837,
-      "step": 45000
-    },
-    {
-      "epoch": 0.65,
-      "learning_rate": 9.62867186972619e-05,
-      "loss": 2.3821,
-      "step": 45500
-    },
-    {
-      "epoch": 0.66,
-      "learning_rate": 9.623479195550917e-05,
-      "loss": 2.3754,
-      "step": 46000
-    },
-    {
-      "epoch": 0.67,
-      "learning_rate": 9.618286521375643e-05,
-      "loss": 2.371,
-      "step": 46500
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 9.613093847200371e-05,
-      "loss": 2.3657,
-      "step": 47000
-    },
-    {
-      "epoch": 0.68,
-      "learning_rate": 9.607901173025097e-05,
-      "loss": 2.3574,
-      "step": 47500
-    },
-    {
-      "epoch": 0.69,
-      "learning_rate": 9.602718884198175e-05,
-      "loss": 2.3635,
-      "step": 48000
-    },
-    {
-      "epoch": 0.7,
-      "learning_rate": 9.5975262100229e-05,
-      "loss": 2.356,
-      "step": 48500
-    },
-    {
-      "epoch": 0.71,
-      "learning_rate": 9.592333535847627e-05,
-      "loss": 2.3499,
-      "step": 49000
-    },
-    {
-      "epoch": 0.71,
-      "learning_rate": 9.587140861672354e-05,
-      "loss": 2.3485,
-      "step": 49500
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 9.581948187497079e-05,
-      "loss": 2.3504,
-      "step": 50000
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 9.576755513321806e-05,
-      "loss": 2.341,
-      "step": 50500
-    },
-    {
-      "epoch": 0.73,
-      "learning_rate": 9.571573224494883e-05,
-      "loss": 2.3404,
-      "step": 51000
-    },
-    {
-      "epoch": 0.74,
-      "learning_rate": 9.56638055031961e-05,
-      "loss": 2.3353,
-      "step": 51500
-    },
-    {
-      "epoch": 0.75,
-      "learning_rate": 9.561187876144336e-05,
-      "loss": 2.3323,
-      "step": 52000
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 9.555995201969062e-05,
-      "loss": 2.3328,
-      "step": 52500
-    },
-    {
-      "epoch": 0.76,
-      "learning_rate": 9.55081291314214e-05,
-      "loss": 2.3304,
-      "step": 53000
-    },
-    {
-      "epoch": 0.77,
-      "learning_rate": 9.545620238966865e-05,
-      "loss": 2.3251,
-      "step": 53500
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 9.540427564791592e-05,
-      "loss": 2.3243,
-      "step": 54000
-    },
-    {
-      "epoch": 0.78,
-      "learning_rate": 9.535234890616319e-05,
-      "loss": 2.3267,
-      "step": 54500
-    },
-    {
-      "epoch": 0.79,
-      "learning_rate": 9.530042216441046e-05,
-      "loss": 2.3152,
-      "step": 55000
-    },
-    {
-      "epoch": 0.8,
-      "learning_rate": 9.524849542265773e-05,
-      "loss": 2.3157,
-      "step": 55500
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 9.519656868090498e-05,
-      "loss": 2.3077,
-      "step": 56000
-    },
-    {
-      "epoch": 0.81,
-      "learning_rate": 9.514474579263576e-05,
-      "loss": 2.3043,
-      "step": 56500
-    },
-    {
-      "epoch": 0.82,
-      "learning_rate": 9.509281905088301e-05,
-      "loss": 2.3081,
-      "step": 57000
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 9.504089230913028e-05,
-      "loss": 2.2992,
-      "step": 57500
-    },
-    {
-      "epoch": 0.83,
-      "learning_rate": 9.498896556737755e-05,
-      "loss": 2.3053,
-      "step": 58000
-    },
-    {
-      "epoch": 0.84,
-      "learning_rate": 9.493714267910832e-05,
-      "loss": 2.2959,
-      "step": 58500
-    },
-    {
-      "epoch": 0.85,
-      "learning_rate": 9.488521593735559e-05,
-      "loss": 2.2958,
-      "step": 59000
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 9.483328919560285e-05,
-      "loss": 2.292,
-      "step": 59500
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 9.478136245385011e-05,
-      "loss": 2.2902,
-      "step": 60000
-    },
-    {
-      "epoch": 0.87,
-      "learning_rate": 9.472943571209738e-05,
-      "loss": 2.2859,
-      "step": 60500
-    },
-    {
-      "epoch": 0.88,
-      "learning_rate": 9.467750897034465e-05,
-      "loss": 2.2821,
-      "step": 61000
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 9.46255822285919e-05,
-      "loss": 2.2826,
-      "step": 61500
-    },
-    {
-      "epoch": 0.89,
-      "learning_rate": 9.457365548683917e-05,
-      "loss": 2.2825,
-      "step": 62000
-    },
-    {
-      "epoch": 0.9,
-      "learning_rate": 9.452183259856993e-05,
-      "loss": 2.2826,
-      "step": 62500
-    },
-    {
-      "epoch": 0.91,
-      "learning_rate": 9.44699058568172e-05,
-      "loss": 2.2723,
-      "step": 63000
-    },
-    {
-      "epoch": 0.91,
-      "learning_rate": 9.441797911506447e-05,
-      "loss": 2.2721,
-      "step": 63500
-    },
-    {
-      "epoch": 0.92,
-      "learning_rate": 9.436605237331174e-05,
-      "loss": 2.2744,
-      "step": 64000
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 9.431412563155901e-05,
-      "loss": 2.2722,
-      "step": 64500
-    },
-    {
-      "epoch": 0.94,
-      "learning_rate": 9.426230274328977e-05,
-      "loss": 2.2649,
-      "step": 65000
-    },
-    {
-      "epoch": 0.94,
-      "learning_rate": 9.421037600153704e-05,
-      "loss": 2.2639,
-      "step": 65500
-    },
-    {
-      "epoch": 0.95,
-      "learning_rate": 9.41584492597843e-05,
-      "loss": 2.2653,
-      "step": 66000
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 9.410652251803157e-05,
-      "loss": 2.2565,
-      "step": 66500
-    },
-    {
-      "epoch": 0.96,
-      "learning_rate": 9.405459577627883e-05,
-      "loss": 2.2624,
-      "step": 67000
-    },
-    {
-      "epoch": 0.97,
-      "learning_rate": 9.400266903452609e-05,
-      "loss": 2.2556,
-      "step": 67500
-    },
-    {
-      "epoch": 0.98,
-      "learning_rate": 9.395074229277337e-05,
-      "loss": 2.2516,
-      "step": 68000
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 9.389881555102063e-05,
-      "loss": 2.249,
-      "step": 68500
-    },
-    {
-      "epoch": 0.99,
-      "learning_rate": 9.38469926627514e-05,
-      "loss": 2.2548,
-      "step": 69000
-    },
-    {
-      "epoch": 1.0,
-      "eval_accuracy": 0.588244433320517,
-      "eval_loss": 2.116262674331665,
-      "eval_runtime": 646.5563,
-      "eval_samples_per_second": 833.55,
-      "eval_steps_per_second": 34.732,
-      "step": 69473
-    },
-    {
-      "epoch": 1.0,
-      "learning_rate": 9.379506592099866e-05,
-      "loss": 2.2492,
-      "step": 69500
-    },
-    {
-      "epoch": 1.01,
-      "learning_rate": 9.374313917924591e-05,
-      "loss": 2.2478,
-      "step": 70000
-    },
-    {
-      "epoch": 1.01,
-      "learning_rate": 9.36912124374932e-05,
-      "loss": 2.2434,
-      "step": 70500
-    },
-    {
-      "epoch": 1.02,
-      "learning_rate": 9.363938954922395e-05,
-      "loss": 2.2369,
-      "step": 71000
-    },
-    {
-      "epoch": 1.03,
-      "learning_rate": 9.358746280747123e-05,
-      "loss": 2.2392,
-      "step": 71500
-    },
-    {
-      "epoch": 1.04,
-      "learning_rate": 9.353553606571849e-05,
-      "loss": 2.2348,
-      "step": 72000
-    },
-    {
-      "epoch": 1.04,
-      "learning_rate": 9.348360932396575e-05,
-      "loss": 2.2382,
-      "step": 72500
-    },
-    {
-      "epoch": 1.05,
-      "learning_rate": 9.343178643569652e-05,
-      "loss": 2.2336,
-      "step": 73000
-    },
-    {
-      "epoch": 1.06,
-      "learning_rate": 9.337985969394379e-05,
-      "loss": 2.2339,
-      "step": 73500
-    },
-    {
-      "epoch": 1.07,
-      "learning_rate": 9.332793295219106e-05,
-      "loss": 2.2286,
-      "step": 74000
-    },
-    {
-      "epoch": 1.07,
-      "learning_rate": 9.327600621043831e-05,
-      "loss": 2.2303,
-      "step": 74500
-    },
-    {
-      "epoch": 1.08,
-      "learning_rate": 9.322418332216909e-05,
-      "loss": 2.2318,
-      "step": 75000
-    },
-    {
-      "epoch": 1.09,
-      "learning_rate": 9.317225658041635e-05,
-      "loss": 2.2249,
-      "step": 75500
-    },
-    {
-      "epoch": 1.09,
-      "learning_rate": 9.312032983866363e-05,
-      "loss": 2.2209,
-      "step": 76000
-    },
-    {
-      "epoch": 1.1,
-      "learning_rate": 9.306840309691088e-05,
-      "loss": 2.223,
-      "step": 76500
-    },
-    {
-      "epoch": 1.11,
-      "learning_rate": 9.301658020864166e-05,
-      "loss": 2.2224,
-      "step": 77000
-    },
-    {
-      "epoch": 1.12,
-      "learning_rate": 9.296465346688892e-05,
-      "loss": 2.2191,
-      "step": 77500
-    },
-    {
-      "epoch": 1.12,
-      "learning_rate": 9.291272672513619e-05,
-      "loss": 2.2184,
-      "step": 78000
-    },
-    {
-      "epoch": 1.13,
-      "learning_rate": 9.286079998338345e-05,
-      "loss": 2.2153,
-      "step": 78500
-    },
-    {
-      "epoch": 1.14,
-      "learning_rate": 9.280897709511422e-05,
-      "loss": 2.212,
-      "step": 79000
-    },
-    {
-      "epoch": 1.14,
-      "learning_rate": 9.275705035336149e-05,
-      "loss": 2.2111,
-      "step": 79500
-    },
-    {
-      "epoch": 1.15,
-      "learning_rate": 9.270512361160874e-05,
-      "loss": 2.2102,
-      "step": 80000
-    },
-    {
-      "epoch": 1.16,
-      "learning_rate": 9.265319686985601e-05,
-      "loss": 2.2068,
-      "step": 80500
-    },
-    {
-      "epoch": 1.17,
-      "learning_rate": 9.260137398158678e-05,
-      "loss": 2.2086,
-      "step": 81000
-    },
-    {
-      "epoch": 1.17,
-      "learning_rate": 9.254944723983405e-05,
-      "loss": 2.2048,
-      "step": 81500
-    },
-    {
-      "epoch": 1.18,
-      "learning_rate": 9.249752049808131e-05,
-      "loss": 2.2054,
-      "step": 82000
-    },
-    {
-      "epoch": 1.19,
-      "learning_rate": 9.244559375632857e-05,
-      "loss": 2.2023,
-      "step": 82500
-    },
-    {
-      "epoch": 1.19,
-      "learning_rate": 9.239377086805935e-05,
-      "loss": 2.2023,
-      "step": 83000
-    },
-    {
-      "epoch": 1.2,
-      "learning_rate": 9.23418441263066e-05,
-      "loss": 2.2016,
-      "step": 83500
-    },
-    {
-      "epoch": 1.21,
-      "learning_rate": 9.228991738455389e-05,
-      "loss": 2.2034,
-      "step": 84000
-    },
-    {
-      "epoch": 1.22,
-      "learning_rate": 9.223799064280114e-05,
-      "loss": 2.1933,
-      "step": 84500
-    },
-    {
-      "epoch": 1.22,
-      "learning_rate": 9.218616775453192e-05,
-      "loss": 2.1979,
-      "step": 85000
-    },
-    {
-      "epoch": 1.23,
-      "learning_rate": 9.213424101277917e-05,
-      "loss": 2.1929,
-      "step": 85500
-    },
-    {
-      "epoch": 1.24,
-      "learning_rate": 9.208231427102644e-05,
-      "loss": 2.1989,
-      "step": 86000
-    },
-    {
-      "epoch": 1.25,
-      "learning_rate": 9.203038752927371e-05,
-      "loss": 2.1941,
-      "step": 86500
-    },
-    {
-      "epoch": 1.25,
-      "learning_rate": 9.197856464100448e-05,
-      "loss": 2.1927,
-      "step": 87000
-    },
-    {
-      "epoch": 1.26,
-      "learning_rate": 9.192663789925175e-05,
-      "loss": 2.1899,
-      "step": 87500
-    },
-    {
-      "epoch": 1.27,
-      "learning_rate": 9.1874711157499e-05,
-      "loss": 2.187,
-      "step": 88000
-    },
-    {
-      "epoch": 1.27,
-      "learning_rate": 9.182278441574627e-05,
-      "loss": 2.1863,
-      "step": 88500
-    },
-    {
-      "epoch": 1.28,
-      "learning_rate": 9.177096152747703e-05,
-      "loss": 2.1881,
-      "step": 89000
-    },
-    {
-      "epoch": 1.29,
-      "learning_rate": 9.17190347857243e-05,
-      "loss": 2.1864,
-      "step": 89500
-    },
-    {
-      "epoch": 1.3,
-      "learning_rate": 9.166710804397157e-05,
-      "loss": 2.1846,
-      "step": 90000
-    },
-    {
-      "epoch": 1.3,
-      "learning_rate": 9.161518130221884e-05,
-      "loss": 2.1793,
-      "step": 90500
-    },
-    {
-      "epoch": 1.31,
-      "learning_rate": 9.15633584139496e-05,
-      "loss": 2.1813,
-      "step": 91000
-    },
-    {
-      "epoch": 1.32,
-      "learning_rate": 9.151143167219687e-05,
-      "loss": 2.1755,
-      "step": 91500
-    },
-    {
-      "epoch": 1.32,
-      "learning_rate": 9.145950493044413e-05,
-      "loss": 2.1763,
-      "step": 92000
-    },
-    {
-      "epoch": 1.33,
-      "learning_rate": 9.14075781886914e-05,
-      "loss": 2.1777,
-      "step": 92500
-    },
-    {
-      "epoch": 1.34,
-      "learning_rate": 9.135575530042216e-05,
-      "loss": 2.1759,
-      "step": 93000
-    },
-    {
-      "epoch": 1.35,
-      "learning_rate": 9.130382855866943e-05,
-      "loss": 2.1758,
-      "step": 93500
-    },
-    {
-      "epoch": 1.35,
-      "learning_rate": 9.12519018169167e-05,
-      "loss": 2.1762,
-      "step": 94000
-    },
-    {
-      "epoch": 1.36,
-      "learning_rate": 9.119997507516397e-05,
-      "loss": 2.1713,
-      "step": 94500
-    },
-    {
-      "epoch": 1.37,
-      "learning_rate": 9.114815218689473e-05,
-      "loss": 2.1713,
-      "step": 95000
-    },
-    {
-      "epoch": 1.37,
-      "learning_rate": 9.1096225445142e-05,
-      "loss": 2.1681,
-      "step": 95500
-    },
-    {
-      "epoch": 1.38,
-      "learning_rate": 9.104429870338926e-05,
-      "loss": 2.1703,
-      "step": 96000
-    },
-    {
-      "epoch": 1.39,
-      "learning_rate": 9.099237196163653e-05,
-      "loss": 2.168,
-      "step": 96500
-    },
-    {
-      "epoch": 1.4,
-      "learning_rate": 9.094054907336729e-05,
-      "loss": 2.1642,
-      "step": 97000
-    },
-    {
-      "epoch": 1.4,
-      "learning_rate": 9.088862233161456e-05,
-      "loss": 2.1629,
-      "step": 97500
-    },
-    {
-      "epoch": 1.41,
-      "learning_rate": 9.083669558986183e-05,
-      "loss": 2.1655,
-      "step": 98000
-    },
-    {
-      "epoch": 1.42,
-      "learning_rate": 9.07847688481091e-05,
-      "loss": 2.1653,
-      "step": 98500
-    },
-    {
-      "epoch": 1.43,
-      "learning_rate": 9.073294595983986e-05,
-      "loss": 2.1619,
-      "step": 99000
-    },
-    {
-      "epoch": 1.43,
-      "learning_rate": 9.068101921808713e-05,
-      "loss": 2.1608,
-      "step": 99500
-    },
-    {
-      "epoch": 1.44,
-      "learning_rate": 9.062909247633439e-05,
-      "loss": 2.1614,
-      "step": 100000
-    },
-    {
-      "epoch": 1.45,
-      "learning_rate": 9.057716573458165e-05,
-      "loss": 2.1624,
-      "step": 100500
-    },
-    {
-      "epoch": 1.45,
-      "learning_rate": 9.052534284631242e-05,
-      "loss": 2.1523,
-      "step": 101000
-    },
-    {
-      "epoch": 1.46,
-      "learning_rate": 9.047341610455969e-05,
-      "loss": 2.1582,
-      "step": 101500
-    },
-    {
-      "epoch": 1.47,
-      "learning_rate": 9.042148936280696e-05,
-      "loss": 2.1586,
-      "step": 102000
-    },
-    {
-      "epoch": 1.48,
-      "learning_rate": 9.036956262105421e-05,
-      "loss": 2.1508,
-      "step": 102500
-    },
-    {
-      "epoch": 1.48,
-      "learning_rate": 9.031773973278499e-05,
-      "loss": 2.1563,
-      "step": 103000
-    },
-    {
-      "epoch": 1.49,
-      "learning_rate": 9.026581299103226e-05,
-      "loss": 2.1538,
-      "step": 103500
-    },
-    {
-      "epoch": 1.5,
-      "learning_rate": 9.021388624927953e-05,
-      "loss": 2.151,
-      "step": 104000
-    },
-    {
-      "epoch": 1.5,
-      "learning_rate": 9.016195950752678e-05,
-      "loss": 2.1447,
-      "step": 104500
-    },
-    {
-      "epoch": 1.51,
-      "learning_rate": 9.011013661925756e-05,
-      "loss": 2.1508,
-      "step": 105000
-    },
-    {
-      "epoch": 1.52,
-      "learning_rate": 9.005820987750482e-05,
-      "loss": 2.145,
-      "step": 105500
-    },
-    {
-      "epoch": 1.53,
-      "learning_rate": 9.000628313575209e-05,
-      "loss": 2.1438,
-      "step": 106000
-    },
-    {
-      "epoch": 1.53,
-      "learning_rate": 8.995435639399935e-05,
-      "loss": 2.1465,
-      "step": 106500
-    },
-    {
-      "epoch": 1.54,
-      "learning_rate": 8.990253350573012e-05,
-      "loss": 2.1456,
-      "step": 107000
-    },
-    {
-      "epoch": 1.55,
-      "learning_rate": 8.985060676397739e-05,
-      "loss": 2.144,
-      "step": 107500
-    },
-    {
-      "epoch": 1.55,
-      "learning_rate": 8.979868002222464e-05,
-      "loss": 2.1466,
-      "step": 108000
-    },
-    {
-      "epoch": 1.56,
-      "learning_rate": 8.974675328047191e-05,
-      "loss": 2.1409,
-      "step": 108500
-    },
-    {
-      "epoch": 1.57,
-      "learning_rate": 8.969493039220268e-05,
-      "loss": 2.1367,
-      "step": 109000
-    },
-    {
-      "epoch": 1.58,
-      "learning_rate": 8.964300365044995e-05,
-      "loss": 2.1402,
-      "step": 109500
-    },
-    {
-      "epoch": 1.58,
-      "learning_rate": 8.959107690869721e-05,
-      "loss": 2.1362,
-      "step": 110000
-    },
-    {
-      "epoch": 1.59,
-      "learning_rate": 8.953915016694447e-05,
-      "loss": 2.1351,
-      "step": 110500
-    },
-    {
-      "epoch": 1.6,
-      "learning_rate": 8.948732727867525e-05,
-      "loss": 2.1379,
-      "step": 111000
-    },
-    {
-      "epoch": 1.6,
-      "learning_rate": 8.94354005369225e-05,
-      "loss": 2.134,
-      "step": 111500
-    },
-    {
-      "epoch": 1.61,
-      "learning_rate": 8.938347379516979e-05,
-      "loss": 2.1356,
-      "step": 112000
-    },
-    {
-      "epoch": 1.62,
-      "learning_rate": 8.933154705341704e-05,
-      "loss": 2.1365,
-      "step": 112500
-    },
-    {
-      "epoch": 1.63,
-      "learning_rate": 8.927972416514782e-05,
-      "loss": 2.1324,
-      "step": 113000
-    },
-    {
-      "epoch": 1.63,
-      "learning_rate": 8.922779742339508e-05,
-      "loss": 2.1308,
-      "step": 113500
-    },
-    {
-      "epoch": 1.64,
-      "learning_rate": 8.917587068164234e-05,
-      "loss": 2.1312,
-      "step": 114000
-    },
-    {
-      "epoch": 1.65,
-      "learning_rate": 8.912394393988961e-05,
-      "loss": 2.1292,
-      "step": 114500
-    },
-    {
-      "epoch": 1.66,
-      "learning_rate": 8.907212105162038e-05,
-      "loss": 2.1316,
-      "step": 115000
-    },
-    {
-      "epoch": 1.66,
-      "learning_rate": 8.902019430986765e-05,
-      "loss": 2.1271,
-      "step": 115500
-    },
-    {
-      "epoch": 1.67,
-      "learning_rate": 8.89682675681149e-05,
-      "loss": 2.128,
-      "step": 116000
-    },
-    {
-      "epoch": 1.68,
-      "learning_rate": 8.891634082636218e-05,
-      "loss": 2.1276,
-      "step": 116500
-    },
-    {
-      "epoch": 1.68,
-      "learning_rate": 8.886451793809294e-05,
-      "loss": 2.1242,
-      "step": 117000
-    },
-    {
-      "epoch": 1.69,
-      "learning_rate": 8.881259119634022e-05,
-      "loss": 2.1264,
-      "step": 117500
-    },
-    {
-      "epoch": 1.7,
-      "learning_rate": 8.876066445458747e-05,
-      "loss": 2.1215,
-      "step": 118000
-    },
-    {
-      "epoch": 1.71,
-      "learning_rate": 8.870873771283473e-05,
-      "loss": 2.1246,
-      "step": 118500
-    },
-    {
-      "epoch": 1.71,
-      "learning_rate": 8.86569148245655e-05,
-      "loss": 2.1247,
-      "step": 119000
-    },
-    {
-      "epoch": 1.72,
-      "learning_rate": 8.860498808281276e-05,
-      "loss": 2.1228,
-      "step": 119500
-    },
-    {
-      "epoch": 1.73,
-      "learning_rate": 8.855306134106004e-05,
-      "loss": 2.1223,
-      "step": 120000
-    },
-    {
-      "epoch": 1.73,
-      "learning_rate": 8.85011345993073e-05,
-      "loss": 2.1202,
-      "step": 120500
-    },
-    {
-      "epoch": 1.74,
-      "learning_rate": 8.844931171103808e-05,
-      "loss": 2.1223,
-      "step": 121000
-    },
-    {
-      "epoch": 1.75,
-      "learning_rate": 8.839738496928533e-05,
-      "loss": 2.1198,
-      "step": 121500
-    },
-    {
-      "epoch": 1.76,
-      "learning_rate": 8.83454582275326e-05,
-      "loss": 2.1164,
-      "step": 122000
-    },
-    {
-      "epoch": 1.76,
-      "learning_rate": 8.829353148577987e-05,
-      "loss": 2.1192,
-      "step": 122500
-    },
-    {
-      "epoch": 1.77,
-      "learning_rate": 8.824170859751064e-05,
-      "loss": 2.1152,
-      "step": 123000
-    },
-    {
-      "epoch": 1.78,
-      "learning_rate": 8.81897818557579e-05,
-      "loss": 2.1163,
-      "step": 123500
-    },
-    {
-      "epoch": 1.78,
-      "learning_rate": 8.813785511400516e-05,
-      "loss": 2.1119,
-      "step": 124000
-    },
-    {
-      "epoch": 1.79,
-      "learning_rate": 8.808592837225244e-05,
-      "loss": 2.113,
-      "step": 124500
-    },
-    {
-      "epoch": 1.8,
-      "learning_rate": 8.803410548398319e-05,
-      "loss": 2.1135,
-      "step": 125000
-    },
-    {
-      "epoch": 1.81,
-      "learning_rate": 8.798217874223047e-05,
-      "loss": 2.1099,
-      "step": 125500
-    },
-    {
-      "epoch": 1.81,
-      "learning_rate": 8.793025200047773e-05,
-      "loss": 2.1139,
-      "step": 126000
-    },
-    {
-      "epoch": 1.82,
-      "learning_rate": 8.7878325258725e-05,
-      "loss": 2.1054,
-      "step": 126500
-    },
-    {
-      "epoch": 1.83,
-      "learning_rate": 8.782650237045576e-05,
-      "loss": 2.1085,
-      "step": 127000
-    },
-    {
-      "epoch": 1.84,
-      "learning_rate": 8.777457562870303e-05,
-      "loss": 2.1115,
-      "step": 127500
-    },
-    {
-      "epoch": 1.84,
-      "learning_rate": 8.77226488869503e-05,
-      "loss": 2.1057,
-      "step": 128000
-    },
-    {
-      "epoch": 1.85,
-      "learning_rate": 8.767072214519756e-05,
-      "loss": 2.1108,
-      "step": 128500
-    },
-    {
-      "epoch": 1.86,
-      "learning_rate": 8.761889925692834e-05,
-      "loss": 2.1105,
-      "step": 129000
-    },
-    {
-      "epoch": 1.86,
-      "learning_rate": 8.756697251517559e-05,
-      "loss": 2.1075,
-      "step": 129500
-    },
-    {
-      "epoch": 1.87,
-      "learning_rate": 8.751504577342286e-05,
-      "loss": 2.1027,
-      "step": 130000
-    },
-    {
-      "epoch": 1.88,
-      "learning_rate": 8.746311903167013e-05,
-      "loss": 2.1043,
-      "step": 130500
-    },
-    {
-      "epoch": 1.89,
-      "learning_rate": 8.741129614340089e-05,
-      "loss": 2.0995,
-      "step": 131000
-    },
-    {
-      "epoch": 1.89,
-      "learning_rate": 8.735936940164816e-05,
-      "loss": 2.1022,
-      "step": 131500
-    },
-    {
-      "epoch": 1.9,
-      "learning_rate": 8.730744265989542e-05,
-      "loss": 2.1011,
-      "step": 132000
-    },
-    {
-      "epoch": 1.91,
-      "learning_rate": 8.72555159181427e-05,
-      "loss": 2.1012,
-      "step": 132500
-    },
-    {
-      "epoch": 1.91,
-      "learning_rate": 8.720369302987345e-05,
-      "loss": 2.1012,
-      "step": 133000
-    },
-    {
-      "epoch": 1.92,
-      "learning_rate": 8.715176628812073e-05,
-      "loss": 2.1026,
-      "step": 133500
-    },
-    {
-      "epoch": 1.93,
-      "learning_rate": 8.709983954636799e-05,
-      "loss": 2.099,
-      "step": 134000
-    },
-    {
-      "epoch": 1.94,
-      "learning_rate": 8.704791280461526e-05,
-      "loss": 2.1023,
-      "step": 134500
-    },
-    {
-      "epoch": 1.94,
-      "learning_rate": 8.699608991634602e-05,
-      "loss": 2.0986,
-      "step": 135000
-    },
-    {
-      "epoch": 1.95,
-      "learning_rate": 8.694416317459329e-05,
-      "loss": 2.0948,
-      "step": 135500
-    },
-    {
-      "epoch": 1.96,
-      "learning_rate": 8.689223643284056e-05,
-      "loss": 2.0944,
-      "step": 136000
-    },
-    {
-      "epoch": 1.96,
-      "learning_rate": 8.684030969108781e-05,
-      "loss": 2.0987,
-      "step": 136500
-    },
-    {
-      "epoch": 1.97,
-      "learning_rate": 8.678848680281859e-05,
-      "loss": 2.0933,
-      "step": 137000
-    },
-    {
-      "epoch": 1.98,
-      "learning_rate": 8.673656006106585e-05,
-      "loss": 2.0973,
-      "step": 137500
-    },
-    {
-      "epoch": 1.99,
-      "learning_rate": 8.668463331931312e-05,
-      "loss": 2.093,
-      "step": 138000
-    },
-    {
-      "epoch": 1.99,
-      "learning_rate": 8.663270657756038e-05,
-      "loss": 2.0904,
-      "step": 138500
-    },
-    {
-      "epoch": 2.0,
-      "eval_accuracy": 0.6101490861863045,
-      "eval_loss": 1.9561994075775146,
-      "eval_runtime": 646.8164,
-      "eval_samples_per_second": 833.215,
-      "eval_steps_per_second": 34.718,
-      "step": 138946
-    },
-    {
-      "epoch": 2.0,
-      "learning_rate": 8.658077983580765e-05,
-      "loss": 2.0869,
-      "step": 139000
-    },
-    {
-      "epoch": 2.01,
-      "learning_rate": 8.652895694753842e-05,
-      "loss": 2.087,
-      "step": 139500
-    },
-    {
-      "epoch": 2.02,
-      "learning_rate": 8.647703020578569e-05,
-      "loss": 2.0856,
-      "step": 140000
-    },
-    {
-      "epoch": 2.02,
-      "learning_rate": 8.642510346403294e-05,
-      "loss": 2.0843,
-      "step": 140500
-    },
-    {
-      "epoch": 2.03,
-      "learning_rate": 8.637317672228021e-05,
-      "loss": 2.086,
-      "step": 141000
-    },
-    {
-      "epoch": 2.04,
-      "learning_rate": 8.632135383401098e-05,
-      "loss": 2.091,
-      "step": 141500
-    },
-    {
-      "epoch": 2.04,
-      "learning_rate": 8.626942709225824e-05,
-      "loss": 2.083,
-      "step": 142000
-    },
-    {
-      "epoch": 2.05,
-      "learning_rate": 8.621750035050551e-05,
-      "loss": 2.0823,
-      "step": 142500
-    },
-    {
-      "epoch": 2.06,
-      "learning_rate": 8.616557360875278e-05,
-      "loss": 2.0834,
-      "step": 143000
-    },
-    {
-      "epoch": 2.07,
-      "learning_rate": 8.611375072048355e-05,
-      "loss": 2.0838,
-      "step": 143500
-    },
-    {
-      "epoch": 2.07,
-      "learning_rate": 8.606182397873082e-05,
-      "loss": 2.0821,
-      "step": 144000
-    },
-    {
-      "epoch": 2.08,
-      "learning_rate": 8.600989723697807e-05,
-      "loss": 2.082,
-      "step": 144500
-    },
-    {
-      "epoch": 2.09,
-      "learning_rate": 8.595797049522534e-05,
-      "loss": 2.082,
-      "step": 145000
-    },
-    {
-      "epoch": 2.09,
-      "learning_rate": 8.59061476069561e-05,
-      "loss": 2.0836,
-      "step": 145500
-    },
-    {
-      "epoch": 2.1,
-      "learning_rate": 8.585422086520337e-05,
-      "loss": 2.0806,
-      "step": 146000
-    },
-    {
-      "epoch": 2.11,
-      "learning_rate": 8.580229412345064e-05,
-      "loss": 2.08,
-      "step": 146500
-    },
-    {
-      "epoch": 2.12,
-      "learning_rate": 8.575036738169791e-05,
-      "loss": 2.0797,
-      "step": 147000
-    },
-    {
-      "epoch": 2.12,
-      "learning_rate": 8.569854449342868e-05,
-      "loss": 2.0777,
-      "step": 147500
-    },
-    {
-      "epoch": 2.13,
-      "learning_rate": 8.564661775167594e-05,
-      "loss": 2.0794,
-      "step": 148000
-    },
-    {
-      "epoch": 2.14,
-      "learning_rate": 8.55946910099232e-05,
-      "loss": 2.078,
-      "step": 148500
-    },
-    {
-      "epoch": 2.14,
-      "learning_rate": 8.554276426817047e-05,
-      "loss": 2.0795,
-      "step": 149000
-    },
-    {
-      "epoch": 2.15,
-      "learning_rate": 8.549094137990123e-05,
-      "loss": 2.0788,
-      "step": 149500
-    },
-    {
-      "epoch": 2.16,
-      "learning_rate": 8.54390146381485e-05,
-      "loss": 2.0767,
-      "step": 150000
-    },
-    {
-      "epoch": 2.17,
-      "learning_rate": 8.538708789639577e-05,
-      "loss": 2.0736,
-      "step": 150500
-    },
-    {
-      "epoch": 2.17,
-      "learning_rate": 8.533516115464302e-05,
-      "loss": 2.0721,
-      "step": 151000
-    },
-    {
-      "epoch": 2.18,
-      "learning_rate": 8.52833382663738e-05,
-      "loss": 2.0705,
-      "step": 151500
-    },
-    {
-      "epoch": 2.19,
-      "learning_rate": 8.523141152462106e-05,
-      "loss": 2.0733,
-      "step": 152000
-    },
-    {
-      "epoch": 2.2,
-      "learning_rate": 8.517948478286834e-05,
-      "loss": 2.0721,
-      "step": 152500
-    },
-    {
-      "epoch": 2.2,
-      "learning_rate": 8.51275580411156e-05,
-      "loss": 2.0699,
-      "step": 153000
-    },
-    {
-      "epoch": 2.21,
-      "learning_rate": 8.507573515284638e-05,
-      "loss": 2.0709,
-      "step": 153500
-    },
-    {
-      "epoch": 2.22,
-      "learning_rate": 8.502380841109363e-05,
-      "loss": 2.0694,
-      "step": 154000
-    },
-    {
-      "epoch": 2.22,
-      "learning_rate": 8.49718816693409e-05,
-      "loss": 2.0686,
-      "step": 154500
-    },
-    {
-      "epoch": 2.23,
-      "learning_rate": 8.491995492758817e-05,
-      "loss": 2.077,
-      "step": 155000
-    },
-    {
-      "epoch": 2.24,
-      "learning_rate": 8.486813203931893e-05,
-      "loss": 2.0667,
-      "step": 155500
-    },
-    {
-      "epoch": 2.25,
-      "learning_rate": 8.48162052975662e-05,
-      "loss": 2.0669,
-      "step": 156000
-    },
-    {
-      "epoch": 2.25,
-      "learning_rate": 8.476427855581346e-05,
-      "loss": 2.0721,
-      "step": 156500
-    },
-    {
-      "epoch": 2.26,
-      "learning_rate": 8.471235181406072e-05,
-      "loss": 2.0668,
-      "step": 157000
-    },
-    {
-      "epoch": 2.27,
-      "learning_rate": 8.466052892579149e-05,
-      "loss": 2.0673,
-      "step": 157500
-    },
-    {
-      "epoch": 2.27,
-      "learning_rate": 8.460860218403876e-05,
-      "loss": 2.0656,
-      "step": 158000
-    },
-    {
-      "epoch": 2.28,
-      "learning_rate": 8.455667544228603e-05,
-      "loss": 2.0647,
-      "step": 158500
-    },
-    {
-      "epoch": 2.29,
-      "learning_rate": 8.450474870053328e-05,
-      "loss": 2.0649,
-      "step": 159000
-    },
-    {
-      "epoch": 2.3,
-      "learning_rate": 8.445292581226406e-05,
-      "loss": 2.0636,
-      "step": 159500
-    },
-    {
-      "epoch": 2.3,
-      "learning_rate": 8.440099907051132e-05,
-      "loss": 2.0653,
-      "step": 160000
-    },
-    {
-      "epoch": 2.31,
-      "learning_rate": 8.43490723287586e-05,
-      "loss": 2.0663,
-      "step": 160500
-    },
-    {
-      "epoch": 2.32,
-      "learning_rate": 8.429714558700585e-05,
-      "loss": 2.0629,
-      "step": 161000
-    },
-    {
-      "epoch": 2.32,
-      "learning_rate": 8.424532269873663e-05,
-      "loss": 2.0617,
-      "step": 161500
-    },
-    {
-      "epoch": 2.33,
-      "learning_rate": 8.419339595698389e-05,
-      "loss": 2.0626,
-      "step": 162000
-    },
-    {
-      "epoch": 2.34,
-      "learning_rate": 8.414146921523116e-05,
-      "loss": 2.0647,
-      "step": 162500
-    },
-    {
-      "epoch": 2.35,
-      "learning_rate": 8.408954247347842e-05,
-      "loss": 2.064,
-      "step": 163000
-    },
-    {
-      "epoch": 2.35,
-      "learning_rate": 8.403771958520919e-05,
-      "loss": 2.0595,
-      "step": 163500
-    },
-    {
-      "epoch": 2.36,
-      "learning_rate": 8.398579284345646e-05,
-      "loss": 2.0597,
-      "step": 164000
-    },
-    {
-      "epoch": 2.37,
-      "learning_rate": 8.393386610170371e-05,
-      "loss": 2.0564,
-      "step": 164500
-    },
-    {
-      "epoch": 2.38,
-      "learning_rate": 8.3881939359951e-05,
-      "loss": 2.0575,
-      "step": 165000
-    },
-    {
-      "epoch": 2.38,
-      "learning_rate": 8.383011647168175e-05,
-      "loss": 2.0587,
-      "step": 165500
-    },
-    {
-      "epoch": 2.39,
-      "learning_rate": 8.377818972992903e-05,
-      "loss": 2.0572,
-      "step": 166000
-    },
-    {
-      "epoch": 2.4,
-      "learning_rate": 8.372626298817628e-05,
-      "loss": 2.0543,
-      "step": 166500
-    },
-    {
-      "epoch": 2.4,
-      "learning_rate": 8.367433624642354e-05,
-      "loss": 2.0566,
-      "step": 167000
-    },
-    {
-      "epoch": 2.41,
-      "learning_rate": 8.362251335815432e-05,
-      "loss": 2.0577,
-      "step": 167500
-    },
-    {
-      "epoch": 2.42,
-      "learning_rate": 8.357058661640157e-05,
-      "loss": 2.053,
-      "step": 168000
-    },
-    {
-      "epoch": 2.43,
-      "learning_rate": 8.351865987464886e-05,
-      "loss": 2.0561,
-      "step": 168500
-    },
-    {
-      "epoch": 2.43,
-      "learning_rate": 8.346673313289611e-05,
-      "loss": 2.0532,
-      "step": 169000
-    },
-    {
-      "epoch": 2.44,
-      "learning_rate": 8.341491024462689e-05,
-      "loss": 2.0553,
-      "step": 169500
-    },
-    {
-      "epoch": 2.45,
-      "learning_rate": 8.336298350287414e-05,
-      "loss": 2.0547,
-      "step": 170000
-    },
-    {
-      "epoch": 2.45,
-      "learning_rate": 8.331105676112141e-05,
-      "loss": 2.0515,
-      "step": 170500
-    },
-    {
-      "epoch": 2.46,
-      "learning_rate": 8.325913001936868e-05,
-      "loss": 2.0486,
-      "step": 171000
-    },
-    {
-      "epoch": 2.47,
-      "learning_rate": 8.320730713109945e-05,
-      "loss": 2.0561,
-      "step": 171500
-    },
-    {
-      "epoch": 2.48,
-      "learning_rate": 8.315538038934672e-05,
-      "loss": 2.0502,
-      "step": 172000
-    },
-    {
-      "epoch": 2.48,
-      "learning_rate": 8.310345364759397e-05,
-      "loss": 2.0525,
-      "step": 172500
-    },
-    {
-      "epoch": 2.49,
-      "learning_rate": 8.305152690584125e-05,
-      "loss": 2.0482,
-      "step": 173000
-    },
-    {
-      "epoch": 2.5,
-      "learning_rate": 8.2999704017572e-05,
-      "loss": 2.0487,
-      "step": 173500
-    },
-    {
-      "epoch": 2.5,
-      "learning_rate": 8.294777727581929e-05,
-      "loss": 2.054,
-      "step": 174000
-    },
-    {
-      "epoch": 2.51,
-      "learning_rate": 8.289585053406654e-05,
-      "loss": 2.0521,
-      "step": 174500
-    },
-    {
-      "epoch": 2.52,
-      "learning_rate": 8.284392379231381e-05,
-      "loss": 2.0493,
-      "step": 175000
-    },
-    {
-      "epoch": 2.53,
-      "learning_rate": 8.279210090404458e-05,
-      "loss": 2.054,
-      "step": 175500
-    },
-    {
-      "epoch": 2.53,
-      "learning_rate": 8.274017416229184e-05,
-      "loss": 2.0516,
-      "step": 176000
-    },
-    {
-      "epoch": 2.54,
-      "learning_rate": 8.268824742053911e-05,
-      "loss": 2.0515,
-      "step": 176500
-    },
-    {
-      "epoch": 2.55,
-      "learning_rate": 8.263632067878637e-05,
-      "loss": 2.044,
-      "step": 177000
-    },
-    {
-      "epoch": 2.55,
-      "learning_rate": 8.258449779051715e-05,
-      "loss": 2.0461,
-      "step": 177500
-    },
-    {
-      "epoch": 2.56,
-      "learning_rate": 8.25325710487644e-05,
-      "loss": 2.0496,
-      "step": 178000
-    },
-    {
-      "epoch": 2.57,
-      "learning_rate": 8.248064430701167e-05,
-      "loss": 2.0456,
-      "step": 178500
-    },
-    {
-      "epoch": 2.58,
-      "learning_rate": 8.242871756525894e-05,
-      "loss": 2.0436,
-      "step": 179000
-    },
-    {
-      "epoch": 2.58,
-      "learning_rate": 8.23768946769897e-05,
-      "loss": 2.0425,
-      "step": 179500
-    },
-    {
-      "epoch": 2.59,
-      "learning_rate": 8.232496793523697e-05,
-      "loss": 2.0476,
-      "step": 180000
-    },
-    {
-      "epoch": 2.6,
-      "learning_rate": 8.227304119348424e-05,
-      "loss": 2.0475,
-      "step": 180500
-    },
-    {
-      "epoch": 2.61,
-      "learning_rate": 8.22211144517315e-05,
-      "loss": 2.0497,
-      "step": 181000
-    },
-    {
-      "epoch": 2.61,
-      "learning_rate": 8.216929156346228e-05,
-      "loss": 2.0468,
-      "step": 181500
-    },
-    {
-      "epoch": 2.62,
-      "learning_rate": 8.211736482170954e-05,
-      "loss": 2.0421,
-      "step": 182000
-    },
-    {
-      "epoch": 2.63,
-      "learning_rate": 8.20654380799568e-05,
-      "loss": 2.0382,
-      "step": 182500
-    },
-    {
-      "epoch": 2.63,
-      "learning_rate": 8.201351133820407e-05,
-      "loss": 2.0407,
-      "step": 183000
-    },
-    {
-      "epoch": 2.64,
-      "learning_rate": 8.196168844993483e-05,
-      "loss": 2.0413,
-      "step": 183500
-    },
-    {
-      "epoch": 2.65,
-      "learning_rate": 8.19097617081821e-05,
-      "loss": 2.0375,
-      "step": 184000
-    },
-    {
-      "epoch": 2.66,
-      "learning_rate": 8.185783496642937e-05,
-      "loss": 2.0414,
-      "step": 184500
-    },
-    {
-      "epoch": 2.66,
-      "learning_rate": 8.180590822467663e-05,
-      "loss": 2.036,
-      "step": 185000
-    },
-    {
-      "epoch": 2.67,
-      "learning_rate": 8.17540853364074e-05,
-      "loss": 2.0376,
-      "step": 185500
-    },
-    {
-      "epoch": 2.68,
-      "learning_rate": 8.170215859465466e-05,
-      "loss": 2.0401,
-      "step": 186000
-    },
-    {
-      "epoch": 2.68,
-      "learning_rate": 8.165023185290193e-05,
-      "loss": 2.0404,
-      "step": 186500
-    },
-    {
-      "epoch": 2.69,
-      "learning_rate": 8.15983051111492e-05,
-      "loss": 2.0387,
-      "step": 187000
-    },
-    {
-      "epoch": 2.7,
-      "learning_rate": 8.154648222287996e-05,
-      "loss": 2.037,
-      "step": 187500
-    },
-    {
-      "epoch": 2.71,
-      "learning_rate": 8.149455548112723e-05,
-      "loss": 2.0411,
-      "step": 188000
-    },
-    {
-      "epoch": 2.71,
-      "learning_rate": 8.14426287393745e-05,
-      "loss": 2.0364,
-      "step": 188500
-    },
-    {
-      "epoch": 2.72,
-      "learning_rate": 8.139070199762175e-05,
-      "loss": 2.0373,
-      "step": 189000
-    },
-    {
-      "epoch": 2.73,
-      "learning_rate": 8.133887910935253e-05,
-      "loss": 2.0338,
-      "step": 189500
-    },
-    {
-      "epoch": 2.73,
-      "learning_rate": 8.128695236759979e-05,
-      "loss": 2.0353,
-      "step": 190000
-    },
-    {
-      "epoch": 2.74,
-      "learning_rate": 8.123502562584706e-05,
-      "loss": 2.0322,
-      "step": 190500
-    },
-    {
-      "epoch": 2.75,
-      "learning_rate": 8.118309888409432e-05,
-      "loss": 2.0317,
-      "step": 191000
-    },
-    {
-      "epoch": 2.76,
-      "learning_rate": 8.113127599582509e-05,
-      "loss": 2.0289,
-      "step": 191500
-    },
-    {
-      "epoch": 2.76,
-      "learning_rate": 8.107934925407236e-05,
-      "loss": 2.0335,
-      "step": 192000
-    },
-    {
-      "epoch": 2.77,
-      "learning_rate": 8.102742251231963e-05,
-      "loss": 2.038,
-      "step": 192500
-    },
-    {
-      "epoch": 2.78,
-      "learning_rate": 8.09754957705669e-05,
-      "loss": 2.0306,
-      "step": 193000
-    },
-    {
-      "epoch": 2.79,
-      "learning_rate": 8.092367288229766e-05,
-      "loss": 2.0329,
-      "step": 193500
-    },
-    {
-      "epoch": 2.79,
-      "learning_rate": 8.087174614054493e-05,
-      "loss": 2.0326,
-      "step": 194000
-    },
-    {
-      "epoch": 2.8,
-      "learning_rate": 8.081981939879219e-05,
-      "loss": 2.0369,
-      "step": 194500
-    },
-    {
-      "epoch": 2.81,
-      "learning_rate": 8.076789265703945e-05,
-      "loss": 2.0309,
-      "step": 195000
-    },
-    {
-      "epoch": 2.81,
-      "learning_rate": 8.071606976877022e-05,
-      "loss": 2.0313,
-      "step": 195500
-    },
-    {
-      "epoch": 2.82,
-      "learning_rate": 8.066414302701749e-05,
-      "loss": 2.0288,
-      "step": 196000
-    },
-    {
-      "epoch": 2.83,
-      "learning_rate": 8.061221628526476e-05,
-      "loss": 2.0297,
-      "step": 196500
-    },
-    {
-      "epoch": 2.84,
-      "learning_rate": 8.056028954351201e-05,
-      "loss": 2.0264,
-      "step": 197000
-    },
-    {
-      "epoch": 2.84,
-      "learning_rate": 8.050846665524279e-05,
-      "loss": 2.0319,
-      "step": 197500
-    },
-    {
-      "epoch": 2.85,
-      "learning_rate": 8.045653991349005e-05,
-      "loss": 2.0314,
-      "step": 198000
-    },
-    {
-      "epoch": 2.86,
-      "learning_rate": 8.040461317173731e-05,
-      "loss": 2.0281,
-      "step": 198500
-    },
-    {
-      "epoch": 2.86,
-      "learning_rate": 8.035268642998458e-05,
-      "loss": 2.0269,
-      "step": 199000
-    },
-    {
-      "epoch": 2.87,
-      "learning_rate": 8.030086354171535e-05,
-      "loss": 2.022,
-      "step": 199500
-    },
-    {
-      "epoch": 2.88,
-      "learning_rate": 8.024893679996262e-05,
-      "loss": 2.032,
-      "step": 200000
-    },
-    {
-      "epoch": 2.89,
-      "learning_rate": 8.019701005820987e-05,
-      "loss": 2.0283,
-      "step": 200500
-    },
-    {
-      "epoch": 2.89,
-      "learning_rate": 8.014508331645715e-05,
-      "loss": 2.0206,
-      "step": 201000
-    },
-    {
-      "epoch": 2.9,
-      "learning_rate": 8.009326042818792e-05,
-      "loss": 2.0196,
-      "step": 201500
-    },
-    {
-      "epoch": 2.91,
-      "learning_rate": 8.004133368643519e-05,
-      "loss": 2.0243,
-      "step": 202000
-    },
-    {
-      "epoch": 2.91,
-      "learning_rate": 7.998951079816595e-05,
-      "loss": 2.0244,
-      "step": 202500
-    },
-    {
-      "epoch": 2.92,
-      "learning_rate": 7.993758405641322e-05,
-      "loss": 2.026,
-      "step": 203000
-    },
-    {
-      "epoch": 2.93,
-      "learning_rate": 7.988565731466048e-05,
-      "loss": 2.0219,
-      "step": 203500
-    },
-    {
-      "epoch": 2.94,
-      "learning_rate": 7.983373057290775e-05,
-      "loss": 2.0271,
-      "step": 204000
-    },
-    {
-      "epoch": 2.94,
-      "learning_rate": 7.978180383115501e-05,
-      "loss": 2.0275,
-      "step": 204500
-    },
-    {
-      "epoch": 2.95,
-      "learning_rate": 7.972987708940227e-05,
-      "loss": 2.0199,
-      "step": 205000
-    },
-    {
-      "epoch": 2.96,
-      "learning_rate": 7.967795034764955e-05,
-      "loss": 2.0202,
-      "step": 205500
-    },
-    {
-      "epoch": 2.97,
-      "learning_rate": 7.96260236058968e-05,
-      "loss": 2.019,
-      "step": 206000
-    },
-    {
-      "epoch": 2.97,
-      "learning_rate": 7.957420071762758e-05,
-      "loss": 2.0214,
-      "step": 206500
-    },
-    {
-      "epoch": 2.98,
-      "learning_rate": 7.952227397587484e-05,
-      "loss": 2.0192,
-      "step": 207000
-    },
-    {
-      "epoch": 2.99,
-      "learning_rate": 7.94703472341221e-05,
-      "loss": 2.0183,
-      "step": 207500
-    },
-    {
-      "epoch": 2.99,
-      "learning_rate": 7.941842049236938e-05,
-      "loss": 2.0203,
-      "step": 208000
-    },
-    {
-      "epoch": 3.0,
-      "eval_accuracy": 0.6207558846474005,
-      "eval_loss": 1.8847733736038208,
-      "eval_runtime": 647.2657,
-      "eval_samples_per_second": 832.636,
-      "eval_steps_per_second": 34.694,
-      "step": 208419
-    },
-    {
-      "epoch": 3.0,
-      "learning_rate": 7.936659760410013e-05,
-      "loss": 2.0154,
-      "step": 208500
-    },
-    {
-      "epoch": 3.01,
-      "learning_rate": 7.931467086234741e-05,
-      "loss": 2.0093,
-      "step": 209000
-    },
-    {
-      "epoch": 3.02,
-      "learning_rate": 7.926274412059467e-05,
-      "loss": 2.0154,
-      "step": 209500
-    },
-    {
-      "epoch": 3.02,
-      "learning_rate": 7.921081737884193e-05,
-      "loss": 2.0148,
-      "step": 210000
-    },
-    {
-      "epoch": 3.03,
-      "learning_rate": 7.91589944905727e-05,
-      "loss": 2.0151,
-      "step": 210500
-    },
-    {
-      "epoch": 3.04,
-      "learning_rate": 7.910706774881997e-05,
-      "loss": 2.0173,
-      "step": 211000
-    },
-    {
-      "epoch": 3.04,
-      "learning_rate": 7.905514100706724e-05,
-      "loss": 2.012,
-      "step": 211500
-    },
-    {
-      "epoch": 3.05,
-      "learning_rate": 7.900321426531449e-05,
-      "loss": 2.0161,
-      "step": 212000
-    },
-    {
-      "epoch": 3.06,
-      "learning_rate": 7.895139137704527e-05,
-      "loss": 2.0142,
-      "step": 212500
-    },
-    {
-      "epoch": 3.07,
-      "learning_rate": 7.889946463529253e-05,
-      "loss": 2.0158,
-      "step": 213000
-    },
-    {
-      "epoch": 3.07,
-      "learning_rate": 7.884753789353981e-05,
-      "loss": 2.0152,
-      "step": 213500
-    },
-    {
-      "epoch": 3.08,
-      "learning_rate": 7.879561115178706e-05,
-      "loss": 2.013,
-      "step": 214000
-    },
-    {
-      "epoch": 3.09,
-      "learning_rate": 7.874378826351784e-05,
-      "loss": 2.0164,
-      "step": 214500
-    },
-    {
-      "epoch": 3.09,
-      "learning_rate": 7.86918615217651e-05,
-      "loss": 2.0166,
-      "step": 215000
-    },
-    {
-      "epoch": 3.1,
-      "learning_rate": 7.863993478001237e-05,
-      "loss": 2.0151,
-      "step": 215500
-    },
-    {
-      "epoch": 3.11,
-      "learning_rate": 7.858800803825963e-05,
-      "loss": 2.0172,
-      "step": 216000
-    },
-    {
-      "epoch": 3.12,
-      "learning_rate": 7.85361851499904e-05,
-      "loss": 2.0083,
-      "step": 216500
-    },
-    {
-      "epoch": 3.12,
-      "learning_rate": 7.848425840823767e-05,
-      "loss": 2.0134,
-      "step": 217000
-    },
-    {
-      "epoch": 3.13,
-      "learning_rate": 7.843233166648492e-05,
-      "loss": 2.0099,
-      "step": 217500
-    },
-    {
-      "epoch": 3.14,
-      "learning_rate": 7.838040492473219e-05,
-      "loss": 2.0139,
-      "step": 218000
-    },
-    {
-      "epoch": 3.15,
-      "learning_rate": 7.832858203646296e-05,
-      "loss": 2.0073,
-      "step": 218500
-    },
-    {
-      "epoch": 3.15,
-      "learning_rate": 7.827665529471023e-05,
-      "loss": 2.0145,
-      "step": 219000
-    },
-    {
-      "epoch": 3.16,
-      "learning_rate": 7.82247285529575e-05,
-      "loss": 2.0108,
-      "step": 219500
-    },
-    {
-      "epoch": 3.17,
-      "learning_rate": 7.817280181120475e-05,
-      "loss": 2.0082,
-      "step": 220000
-    },
-    {
-      "epoch": 3.17,
-      "learning_rate": 7.812097892293553e-05,
-      "loss": 2.0091,
-      "step": 220500
-    },
-    {
-      "epoch": 3.18,
-      "learning_rate": 7.806905218118278e-05,
-      "loss": 2.0099,
-      "step": 221000
-    },
-    {
-      "epoch": 3.19,
-      "learning_rate": 7.801712543943007e-05,
-      "loss": 2.008,
-      "step": 221500
-    },
-    {
-      "epoch": 3.2,
-      "learning_rate": 7.796519869767732e-05,
-      "loss": 2.0075,
-      "step": 222000
-    },
-    {
-      "epoch": 3.2,
-      "learning_rate": 7.79133758094081e-05,
-      "loss": 2.0079,
-      "step": 222500
-    },
-    {
-      "epoch": 3.21,
-      "learning_rate": 7.786144906765535e-05,
-      "loss": 2.008,
-      "step": 223000
-    },
-    {
-      "epoch": 3.22,
-      "learning_rate": 7.780952232590262e-05,
-      "loss": 2.0103,
-      "step": 223500
-    },
-    {
-      "epoch": 3.22,
-      "learning_rate": 7.775759558414989e-05,
-      "loss": 2.0076,
-      "step": 224000
-    },
-    {
-      "epoch": 3.23,
-      "learning_rate": 7.770577269588066e-05,
-      "loss": 2.0104,
-      "step": 224500
-    },
-    {
-      "epoch": 3.24,
-      "learning_rate": 7.765384595412793e-05,
-      "loss": 2.0062,
-      "step": 225000
-    },
-    {
-      "epoch": 3.25,
-      "learning_rate": 7.760191921237518e-05,
-      "loss": 2.0049,
-      "step": 225500
-    },
-    {
-      "epoch": 3.25,
-      "learning_rate": 7.754999247062245e-05,
-      "loss": 2.0051,
-      "step": 226000
-    },
-    {
-      "epoch": 3.26,
-      "learning_rate": 7.749816958235321e-05,
-      "loss": 2.0002,
-      "step": 226500
-    },
-    {
-      "epoch": 3.27,
-      "learning_rate": 7.744624284060048e-05,
-      "loss": 2.0056,
-      "step": 227000
-    },
-    {
-      "epoch": 3.27,
-      "learning_rate": 7.739431609884775e-05,
-      "loss": 2.0077,
-      "step": 227500
-    },
-    {
-      "epoch": 3.28,
-      "learning_rate": 7.734238935709502e-05,
-      "loss": 2.0082,
-      "step": 228000
-    },
-    {
-      "epoch": 3.29,
-      "learning_rate": 7.729056646882579e-05,
-      "loss": 2.0026,
-      "step": 228500
-    },
-    {
-      "epoch": 3.3,
-      "learning_rate": 7.723863972707305e-05,
-      "loss": 2.0034,
-      "step": 229000
-    },
-    {
-      "epoch": 3.3,
-      "learning_rate": 7.718671298532031e-05,
-      "loss": 2.0043,
-      "step": 229500
-    },
-    {
-      "epoch": 3.31,
-      "learning_rate": 7.713478624356758e-05,
-      "loss": 2.0033,
-      "step": 230000
-    },
-    {
-      "epoch": 3.32,
-      "learning_rate": 7.708296335529836e-05,
-      "loss": 2.0018,
-      "step": 230500
-    },
-    {
-      "epoch": 3.33,
-      "learning_rate": 7.703103661354561e-05,
-      "loss": 2.0024,
-      "step": 231000
-    },
-    {
-      "epoch": 3.33,
-      "learning_rate": 7.697910987179288e-05,
-      "loss": 2.0036,
-      "step": 231500
-    },
-    {
-      "epoch": 3.34,
-      "learning_rate": 7.692718313004015e-05,
-      "loss": 1.9985,
-      "step": 232000
-    },
-    {
-      "epoch": 3.35,
-      "learning_rate": 7.687536024177091e-05,
-      "loss": 2.0016,
-      "step": 232500
-    },
-    {
-      "epoch": 3.35,
-      "learning_rate": 7.682343350001818e-05,
-      "loss": 2.0027,
-      "step": 233000
-    },
-    {
-      "epoch": 3.36,
-      "learning_rate": 7.677150675826544e-05,
-      "loss": 2.0009,
-      "step": 233500
-    },
-    {
-      "epoch": 3.37,
-      "learning_rate": 7.67195800165127e-05,
-      "loss": 2.001,
-      "step": 234000
-    },
-    {
-      "epoch": 3.38,
-      "learning_rate": 7.666775712824347e-05,
-      "loss": 2.0001,
-      "step": 234500
-    },
-    {
-      "epoch": 3.38,
-      "learning_rate": 7.661583038649074e-05,
-      "loss": 2.0002,
-      "step": 235000
-    },
-    {
-      "epoch": 3.39,
-      "learning_rate": 7.656390364473801e-05,
-      "loss": 2.0023,
-      "step": 235500
-    },
-    {
-      "epoch": 3.4,
-      "learning_rate": 7.651197690298528e-05,
-      "loss": 1.9976,
-      "step": 236000
-    },
-    {
-      "epoch": 3.4,
-      "learning_rate": 7.646015401471604e-05,
-      "loss": 1.9986,
-      "step": 236500
-    },
-    {
-      "epoch": 3.41,
-      "learning_rate": 7.640822727296331e-05,
-      "loss": 1.9999,
-      "step": 237000
-    },
-    {
-      "epoch": 3.42,
-      "learning_rate": 7.635630053121057e-05,
-      "loss": 1.998,
-      "step": 237500
-    },
-    {
-      "epoch": 3.43,
-      "learning_rate": 7.630437378945783e-05,
-      "loss": 1.9966,
-      "step": 238000
-    },
-    {
-      "epoch": 3.43,
-      "learning_rate": 7.62525509011886e-05,
-      "loss": 1.9962,
-      "step": 238500
-    },
-    {
-      "epoch": 3.44,
-      "learning_rate": 7.620062415943587e-05,
-      "loss": 1.9994,
-      "step": 239000
-    },
-    {
-      "epoch": 3.45,
-      "learning_rate": 7.614869741768314e-05,
-      "loss": 1.998,
-      "step": 239500
-    },
-    {
-      "epoch": 3.45,
-      "learning_rate": 7.609677067593039e-05,
-      "loss": 1.9962,
-      "step": 240000
-    },
-    {
-      "epoch": 3.46,
-      "learning_rate": 7.604484393417767e-05,
-      "loss": 2.0007,
-      "step": 240500
-    },
-    {
-      "epoch": 3.47,
-      "learning_rate": 7.599291719242493e-05,
-      "loss": 1.9939,
-      "step": 241000
-    },
-    {
-      "epoch": 3.48,
-      "learning_rate": 7.594109430415571e-05,
-      "loss": 1.9966,
-      "step": 241500
-    },
-    {
-      "epoch": 3.48,
-      "learning_rate": 7.588916756240296e-05,
-      "loss": 1.9956,
-      "step": 242000
-    },
-    {
-      "epoch": 3.49,
-      "learning_rate": 7.583724082065023e-05,
-      "loss": 1.99,
-      "step": 242500
-    },
-    {
-      "epoch": 3.5,
-      "learning_rate": 7.57853140788975e-05,
-      "loss": 1.9943,
-      "step": 243000
-    },
-    {
-      "epoch": 3.5,
-      "learning_rate": 7.573338733714475e-05,
-      "loss": 1.9931,
-      "step": 243500
-    },
-    {
-      "epoch": 3.51,
-      "learning_rate": 7.568146059539202e-05,
-      "loss": 1.9946,
-      "step": 244000
-    },
-    {
-      "epoch": 3.52,
-      "learning_rate": 7.562953385363929e-05,
-      "loss": 1.9954,
-      "step": 244500
-    },
-    {
-      "epoch": 3.53,
-      "learning_rate": 7.557760711188655e-05,
-      "loss": 1.9956,
-      "step": 245000
-    },
-    {
-      "epoch": 3.53,
-      "learning_rate": 7.552578422361733e-05,
-      "loss": 1.9904,
-      "step": 245500
-    },
-    {
-      "epoch": 3.54,
-      "learning_rate": 7.547385748186458e-05,
-      "loss": 1.9903,
-      "step": 246000
-    },
-    {
-      "epoch": 3.55,
-      "learning_rate": 7.542193074011186e-05,
-      "loss": 1.9946,
-      "step": 246500
-    },
-    {
-      "epoch": 3.56,
-      "learning_rate": 7.537000399835912e-05,
-      "loss": 1.9943,
-      "step": 247000
-    },
-    {
-      "epoch": 3.56,
-      "learning_rate": 7.53181811100899e-05,
-      "loss": 1.9916,
-      "step": 247500
-    },
-    {
-      "epoch": 3.57,
-      "learning_rate": 7.526625436833715e-05,
-      "loss": 1.9965,
-      "step": 248000
-    },
-    {
-      "epoch": 3.58,
-      "learning_rate": 7.521432762658442e-05,
-      "loss": 1.9904,
-      "step": 248500
-    },
-    {
-      "epoch": 3.58,
-      "learning_rate": 7.516240088483169e-05,
-      "loss": 1.9919,
-      "step": 249000
-    },
-    {
-      "epoch": 3.59,
-      "learning_rate": 7.511057799656245e-05,
-      "loss": 1.9945,
-      "step": 249500
-    },
-    {
-      "epoch": 3.6,
-      "learning_rate": 7.505865125480972e-05,
-      "loss": 1.9884,
-      "step": 250000
-    },
-    {
-      "epoch": 3.61,
-      "learning_rate": 7.500672451305698e-05,
-      "loss": 1.9946,
-      "step": 250500
-    },
-    {
-      "epoch": 3.61,
-      "learning_rate": 7.495479777130426e-05,
-      "loss": 1.9906,
-      "step": 251000
-    },
-    {
-      "epoch": 3.62,
-      "learning_rate": 7.490297488303501e-05,
-      "loss": 1.9934,
-      "step": 251500
-    },
-    {
-      "epoch": 3.63,
-      "learning_rate": 7.48510481412823e-05,
-      "loss": 1.9926,
-      "step": 252000
-    },
-    {
-      "epoch": 3.63,
-      "learning_rate": 7.479922525301305e-05,
-      "loss": 1.9896,
-      "step": 252500
-    },
-    {
-      "epoch": 3.64,
-      "learning_rate": 7.474729851126033e-05,
-      "loss": 1.9932,
-      "step": 253000
-    },
-    {
-      "epoch": 3.65,
-      "learning_rate": 7.469537176950758e-05,
-      "loss": 1.9814,
-      "step": 253500
-    },
-    {
-      "epoch": 3.66,
-      "learning_rate": 7.464344502775484e-05,
-      "loss": 1.9863,
-      "step": 254000
-    },
-    {
-      "epoch": 3.66,
-      "learning_rate": 7.459151828600212e-05,
-      "loss": 1.9919,
-      "step": 254500
-    },
-    {
-      "epoch": 3.67,
-      "learning_rate": 7.453959154424937e-05,
-      "loss": 1.9891,
-      "step": 255000
-    },
-    {
-      "epoch": 3.68,
-      "learning_rate": 7.448766480249664e-05,
-      "loss": 1.9859,
-      "step": 255500
-    },
-    {
-      "epoch": 3.68,
-      "learning_rate": 7.443573806074391e-05,
-      "loss": 1.9864,
-      "step": 256000
-    },
-    {
-      "epoch": 3.69,
-      "learning_rate": 7.438391517247468e-05,
-      "loss": 1.9892,
-      "step": 256500
-    },
-    {
-      "epoch": 3.7,
-      "learning_rate": 7.433198843072195e-05,
-      "loss": 1.9883,
-      "step": 257000
-    },
-    {
-      "epoch": 3.71,
-      "learning_rate": 7.42800616889692e-05,
-      "loss": 1.9887,
-      "step": 257500
-    },
-    {
-      "epoch": 3.71,
-      "learning_rate": 7.422813494721647e-05,
-      "loss": 1.9838,
-      "step": 258000
-    },
-    {
-      "epoch": 3.72,
-      "learning_rate": 7.417631205894723e-05,
-      "loss": 1.985,
-      "step": 258500
-    },
-    {
-      "epoch": 3.73,
-      "learning_rate": 7.41243853171945e-05,
-      "loss": 1.9832,
-      "step": 259000
-    },
-    {
-      "epoch": 3.74,
-      "learning_rate": 7.407245857544177e-05,
-      "loss": 1.9863,
-      "step": 259500
-    },
-    {
-      "epoch": 3.74,
-      "learning_rate": 7.402053183368904e-05,
-      "loss": 1.9831,
-      "step": 260000
-    },
-    {
-      "epoch": 3.75,
-      "learning_rate": 7.39687089454198e-05,
-      "loss": 1.9848,
-      "step": 260500
-    },
-    {
-      "epoch": 3.76,
-      "learning_rate": 7.391678220366707e-05,
-      "loss": 1.9823,
-      "step": 261000
-    },
-    {
-      "epoch": 3.76,
-      "learning_rate": 7.386485546191434e-05,
-      "loss": 1.9822,
-      "step": 261500
-    },
-    {
-      "epoch": 3.77,
-      "learning_rate": 7.38129287201616e-05,
-      "loss": 1.9799,
-      "step": 262000
-    },
-    {
-      "epoch": 3.78,
-      "learning_rate": 7.376110583189238e-05,
-      "loss": 1.9854,
-      "step": 262500
-    },
-    {
-      "epoch": 3.79,
-      "learning_rate": 7.370917909013963e-05,
-      "loss": 1.9844,
-      "step": 263000
-    },
-    {
-      "epoch": 3.79,
-      "learning_rate": 7.36572523483869e-05,
-      "loss": 1.9802,
-      "step": 263500
-    },
-    {
-      "epoch": 3.8,
-      "learning_rate": 7.360532560663417e-05,
-      "loss": 1.9807,
-      "step": 264000
-    },
-    {
-      "epoch": 3.81,
-      "learning_rate": 7.355350271836493e-05,
-      "loss": 1.981,
-      "step": 264500
-    },
-    {
-      "epoch": 3.81,
-      "learning_rate": 7.35015759766122e-05,
-      "loss": 1.9839,
-      "step": 265000
-    },
-    {
-      "epoch": 3.82,
-      "learning_rate": 7.344964923485946e-05,
-      "loss": 1.979,
-      "step": 265500
-    },
-    {
-      "epoch": 3.83,
-      "learning_rate": 7.339772249310673e-05,
-      "loss": 1.9787,
-      "step": 266000
-    },
-    {
-      "epoch": 3.84,
-      "learning_rate": 7.334589960483749e-05,
-      "loss": 1.9799,
-      "step": 266500
-    },
-    {
-      "epoch": 3.84,
-      "learning_rate": 7.329397286308476e-05,
-      "loss": 1.9795,
-      "step": 267000
-    },
-    {
-      "epoch": 3.85,
-      "learning_rate": 7.324204612133203e-05,
-      "loss": 1.9833,
-      "step": 267500
-    },
-    {
-      "epoch": 3.86,
-      "learning_rate": 7.31901193795793e-05,
-      "loss": 1.9806,
-      "step": 268000
-    },
-    {
-      "epoch": 3.86,
-      "learning_rate": 7.313829649131006e-05,
-      "loss": 1.9768,
-      "step": 268500
-    },
-    {
-      "epoch": 3.87,
-      "learning_rate": 7.308636974955733e-05,
-      "loss": 1.9801,
-      "step": 269000
-    },
-    {
-      "epoch": 3.88,
-      "learning_rate": 7.30344430078046e-05,
-      "loss": 1.9756,
-      "step": 269500
-    },
-    {
-      "epoch": 3.89,
-      "learning_rate": 7.298251626605185e-05,
-      "loss": 1.9831,
-      "step": 270000
-    },
-    {
-      "epoch": 3.89,
-      "learning_rate": 7.293069337778263e-05,
-      "loss": 1.9789,
-      "step": 270500
-    },
-    {
-      "epoch": 3.9,
-      "learning_rate": 7.287876663602989e-05,
-      "loss": 1.9765,
-      "step": 271000
-    },
-    {
-      "epoch": 3.91,
-      "learning_rate": 7.282683989427716e-05,
-      "loss": 1.976,
-      "step": 271500
-    },
-    {
-      "epoch": 3.92,
-      "learning_rate": 7.277491315252443e-05,
-      "loss": 1.9803,
-      "step": 272000
-    },
-    {
-      "epoch": 3.92,
-      "learning_rate": 7.27229864107717e-05,
-      "loss": 1.9808,
-      "step": 272500
-    },
-    {
-      "epoch": 3.93,
-      "learning_rate": 7.267105966901895e-05,
-      "loss": 1.9821,
-      "step": 273000
-    },
-    {
-      "epoch": 3.94,
-      "learning_rate": 7.261913292726622e-05,
-      "loss": 1.9778,
-      "step": 273500
-    },
-    {
-      "epoch": 3.94,
-      "learning_rate": 7.256720618551349e-05,
-      "loss": 1.9769,
-      "step": 274000
-    },
-    {
-      "epoch": 3.95,
-      "learning_rate": 7.251538329724425e-05,
-      "loss": 1.9774,
-      "step": 274500
-    },
-    {
-      "epoch": 3.96,
-      "learning_rate": 7.246345655549152e-05,
-      "loss": 1.9731,
-      "step": 275000
-    },
-    {
-      "epoch": 3.97,
-      "learning_rate": 7.241152981373877e-05,
-      "loss": 1.9745,
-      "step": 275500
-    },
-    {
-      "epoch": 3.97,
-      "learning_rate": 7.235960307198604e-05,
-      "loss": 1.979,
-      "step": 276000
-    },
-    {
-      "epoch": 3.98,
-      "learning_rate": 7.230778018371681e-05,
-      "loss": 1.9733,
-      "step": 276500
-    },
-    {
-      "epoch": 3.99,
-      "learning_rate": 7.225585344196408e-05,
-      "loss": 1.9762,
-      "step": 277000
-    },
-    {
-      "epoch": 3.99,
-      "learning_rate": 7.220392670021135e-05,
-      "loss": 1.978,
-      "step": 277500
-    },
-    {
-      "epoch": 4.0,
-      "eval_accuracy": 0.6271896670934867,
-      "eval_loss": 1.8407700061798096,
-      "eval_runtime": 646.5128,
-      "eval_samples_per_second": 833.606,
-      "eval_steps_per_second": 34.734,
-      "step": 277892
-    },
-    {
-      "epoch": 4.0,
-      "learning_rate": 7.21519999584586e-05,
-      "loss": 1.9736,
-      "step": 278000
-    },
-    {
-      "epoch": 4.01,
-      "learning_rate": 7.210017707018938e-05,
-      "loss": 1.9682,
-      "step": 278500
-    },
-    {
-      "epoch": 4.02,
-      "learning_rate": 7.204825032843665e-05,
-      "loss": 1.9706,
-      "step": 279000
-    },
-    {
-      "epoch": 4.02,
-      "learning_rate": 7.199632358668392e-05,
-      "loss": 1.9751,
-      "step": 279500
-    },
-    {
-      "epoch": 4.03,
-      "learning_rate": 7.194439684493117e-05,
-      "loss": 1.9709,
-      "step": 280000
-    },
-    {
-      "epoch": 4.04,
-      "learning_rate": 7.189257395666195e-05,
-      "loss": 1.9697,
-      "step": 280500
-    },
-    {
-      "epoch": 4.04,
-      "learning_rate": 7.18406472149092e-05,
-      "loss": 1.9671,
-      "step": 281000
-    },
-    {
-      "epoch": 4.05,
-      "learning_rate": 7.178872047315647e-05,
-      "loss": 1.9717,
-      "step": 281500
-    },
-    {
-      "epoch": 4.06,
-      "learning_rate": 7.173679373140374e-05,
-      "loss": 1.9722,
-      "step": 282000
-    },
-    {
-      "epoch": 4.07,
-      "learning_rate": 7.168497084313451e-05,
-      "loss": 1.9695,
-      "step": 282500
-    },
-    {
-      "epoch": 4.07,
-      "learning_rate": 7.163314795486528e-05,
-      "loss": 1.9683,
-      "step": 283000
-    },
-    {
-      "epoch": 4.08,
-      "learning_rate": 7.158122121311254e-05,
-      "loss": 1.9647,
-      "step": 283500
-    },
-    {
-      "epoch": 4.09,
-      "learning_rate": 7.152929447135981e-05,
-      "loss": 1.9672,
-      "step": 284000
-    },
-    {
-      "epoch": 4.1,
-      "learning_rate": 7.147736772960707e-05,
-      "loss": 1.9762,
-      "step": 284500
-    },
-    {
-      "epoch": 4.1,
-      "learning_rate": 7.142544098785435e-05,
-      "loss": 1.9708,
-      "step": 285000
-    },
-    {
-      "epoch": 4.11,
-      "learning_rate": 7.13735142461016e-05,
-      "loss": 1.9703,
-      "step": 285500
-    },
-    {
-      "epoch": 4.12,
-      "learning_rate": 7.132158750434886e-05,
-      "loss": 1.9673,
-      "step": 286000
-    },
-    {
-      "epoch": 4.12,
-      "learning_rate": 7.126966076259614e-05,
-      "loss": 1.9704,
-      "step": 286500
-    },
-    {
-      "epoch": 4.13,
-      "learning_rate": 7.121783787432689e-05,
-      "loss": 1.9691,
-      "step": 287000
-    },
-    {
-      "epoch": 4.14,
-      "learning_rate": 7.116591113257417e-05,
-      "loss": 1.9683,
-      "step": 287500
-    },
-    {
-      "epoch": 4.15,
-      "learning_rate": 7.111398439082143e-05,
-      "loss": 1.9687,
-      "step": 288000
-    },
-    {
-      "epoch": 4.15,
-      "learning_rate": 7.10620576490687e-05,
-      "loss": 1.9648,
-      "step": 288500
-    },
-    {
-      "epoch": 4.16,
-      "learning_rate": 7.101023476079946e-05,
-      "loss": 1.9641,
-      "step": 289000
-    },
-    {
-      "epoch": 4.17,
-      "learning_rate": 7.095830801904673e-05,
-      "loss": 1.9635,
-      "step": 289500
-    },
-    {
-      "epoch": 4.17,
-      "learning_rate": 7.0906381277294e-05,
-      "loss": 1.9626,
-      "step": 290000
-    },
-    {
-      "epoch": 4.18,
-      "learning_rate": 7.085445453554125e-05,
-      "loss": 1.9643,
-      "step": 290500
-    },
-    {
-      "epoch": 4.19,
-      "learning_rate": 7.080263164727203e-05,
-      "loss": 1.9661,
-      "step": 291000
-    },
-    {
-      "epoch": 4.2,
-      "learning_rate": 7.075070490551929e-05,
-      "loss": 1.9664,
-      "step": 291500
-    },
-    {
-      "epoch": 4.2,
-      "learning_rate": 7.069877816376657e-05,
-      "loss": 1.9648,
-      "step": 292000
-    },
-    {
-      "epoch": 4.21,
-      "learning_rate": 7.064685142201383e-05,
-      "loss": 1.9605,
-      "step": 292500
-    },
-    {
-      "epoch": 4.22,
-      "learning_rate": 7.05950285337446e-05,
-      "loss": 1.9626,
-      "step": 293000
-    },
-    {
-      "epoch": 4.22,
-      "learning_rate": 7.054310179199186e-05,
-      "loss": 1.9673,
-      "step": 293500
-    },
-    {
-      "epoch": 4.23,
-      "learning_rate": 7.049117505023913e-05,
-      "loss": 1.9646,
-      "step": 294000
-    },
-    {
-      "epoch": 4.24,
-      "learning_rate": 7.04392483084864e-05,
-      "loss": 1.9676,
-      "step": 294500
-    },
-    {
-      "epoch": 4.25,
-      "learning_rate": 7.038742542021716e-05,
-      "loss": 1.9631,
-      "step": 295000
-    },
-    {
-      "epoch": 4.25,
-      "learning_rate": 7.033549867846443e-05,
-      "loss": 1.9696,
-      "step": 295500
-    },
-    {
-      "epoch": 4.26,
-      "learning_rate": 7.028357193671169e-05,
-      "loss": 1.9626,
-      "step": 296000
-    },
-    {
-      "epoch": 4.27,
-      "learning_rate": 7.023164519495895e-05,
-      "loss": 1.964,
-      "step": 296500
-    },
-    {
-      "epoch": 4.28,
-      "learning_rate": 7.017982230668972e-05,
-      "loss": 1.9593,
-      "step": 297000
-    },
-    {
-      "epoch": 4.28,
-      "learning_rate": 7.012789556493699e-05,
-      "loss": 1.966,
-      "step": 297500
-    },
-    {
-      "epoch": 4.29,
-      "learning_rate": 7.007596882318426e-05,
-      "loss": 1.9676,
-      "step": 298000
-    },
-    {
-      "epoch": 4.3,
-      "learning_rate": 7.002404208143151e-05,
-      "loss": 1.9667,
-      "step": 298500
-    },
-    {
-      "epoch": 4.3,
-      "learning_rate": 6.997221919316229e-05,
-      "loss": 1.9589,
-      "step": 299000
-    },
-    {
-      "epoch": 4.31,
-      "learning_rate": 6.992029245140955e-05,
-      "loss": 1.9635,
-      "step": 299500
-    },
-    {
-      "epoch": 4.32,
-      "learning_rate": 6.986836570965683e-05,
-      "loss": 1.9628,
-      "step": 300000
-    },
-    {
-      "epoch": 4.33,
-      "learning_rate": 6.981643896790408e-05,
-      "loss": 1.9599,
-      "step": 300500
-    },
-    {
-      "epoch": 4.33,
-      "learning_rate": 6.976461607963486e-05,
-      "loss": 1.9599,
-      "step": 301000
-    },
-    {
-      "epoch": 4.34,
-      "learning_rate": 6.971268933788212e-05,
-      "loss": 1.96,
-      "step": 301500
-    },
-    {
-      "epoch": 4.35,
-      "learning_rate": 6.966076259612939e-05,
-      "loss": 1.964,
-      "step": 302000
-    },
-    {
-      "epoch": 4.35,
-      "learning_rate": 6.960883585437665e-05,
-      "loss": 1.9602,
-      "step": 302500
-    },
-    {
-      "epoch": 4.36,
-      "learning_rate": 6.955701296610742e-05,
-      "loss": 1.9607,
-      "step": 303000
-    },
-    {
-      "epoch": 4.37,
-      "learning_rate": 6.950508622435469e-05,
-      "loss": 1.96,
-      "step": 303500
-    },
-    {
-      "epoch": 4.38,
-      "learning_rate": 6.945315948260194e-05,
-      "loss": 1.9582,
-      "step": 304000
-    },
-    {
-      "epoch": 4.38,
-      "learning_rate": 6.940123274084921e-05,
-      "loss": 1.9645,
-      "step": 304500
-    },
-    {
-      "epoch": 4.39,
-      "learning_rate": 6.934940985257998e-05,
-      "loss": 1.9649,
-      "step": 305000
-    },
-    {
-      "epoch": 4.4,
-      "learning_rate": 6.929748311082725e-05,
-      "loss": 1.9614,
-      "step": 305500
-    },
-    {
-      "epoch": 4.4,
-      "learning_rate": 6.924555636907451e-05,
-      "loss": 1.9581,
-      "step": 306000
-    },
-    {
-      "epoch": 4.41,
-      "learning_rate": 6.919362962732178e-05,
-      "loss": 1.9605,
-      "step": 306500
-    },
-    {
-      "epoch": 4.42,
-      "learning_rate": 6.914180673905255e-05,
-      "loss": 1.9589,
-      "step": 307000
-    },
-    {
-      "epoch": 4.43,
-      "learning_rate": 6.908987999729982e-05,
-      "loss": 1.9598,
-      "step": 307500
-    },
-    {
-      "epoch": 4.43,
-      "learning_rate": 6.903795325554707e-05,
-      "loss": 1.9611,
-      "step": 308000
-    },
-    {
-      "epoch": 4.44,
-      "learning_rate": 6.898602651379434e-05,
-      "loss": 1.9588,
-      "step": 308500
-    },
-    {
-      "epoch": 4.45,
-      "learning_rate": 6.893420362552512e-05,
-      "loss": 1.9596,
-      "step": 309000
-    },
-    {
-      "epoch": 4.45,
-      "learning_rate": 6.888227688377238e-05,
-      "loss": 1.9542,
-      "step": 309500
-    },
-    {
-      "epoch": 4.46,
-      "learning_rate": 6.883035014201964e-05,
-      "loss": 1.9565,
-      "step": 310000
-    },
-    {
-      "epoch": 4.47,
-      "learning_rate": 6.877842340026691e-05,
-      "loss": 1.9596,
-      "step": 310500
-    },
-    {
-      "epoch": 4.48,
-      "learning_rate": 6.872660051199768e-05,
-      "loss": 1.9577,
-      "step": 311000
-    },
-    {
-      "epoch": 4.48,
-      "learning_rate": 6.867467377024495e-05,
-      "loss": 1.9559,
-      "step": 311500
-    },
-    {
-      "epoch": 4.49,
-      "learning_rate": 6.86227470284922e-05,
-      "loss": 1.962,
-      "step": 312000
-    },
-    {
-      "epoch": 4.5,
-      "learning_rate": 6.857082028673947e-05,
-      "loss": 1.958,
-      "step": 312500
-    },
-    {
-      "epoch": 4.51,
-      "learning_rate": 6.851899739847024e-05,
-      "loss": 1.9559,
-      "step": 313000
-    },
-    {
-      "epoch": 4.51,
-      "learning_rate": 6.84670706567175e-05,
-      "loss": 1.9577,
-      "step": 313500
-    },
-    {
-      "epoch": 4.52,
-      "learning_rate": 6.841514391496477e-05,
-      "loss": 1.957,
-      "step": 314000
-    },
-    {
-      "epoch": 4.53,
-      "learning_rate": 6.836321717321204e-05,
-      "loss": 1.9549,
-      "step": 314500
-    },
-    {
-      "epoch": 4.53,
-      "learning_rate": 6.83113942849428e-05,
-      "loss": 1.9533,
-      "step": 315000
-    },
-    {
-      "epoch": 4.54,
-      "learning_rate": 6.825946754319007e-05,
-      "loss": 1.9567,
-      "step": 315500
-    },
-    {
-      "epoch": 4.55,
-      "learning_rate": 6.820754080143733e-05,
-      "loss": 1.9572,
-      "step": 316000
-    },
-    {
-      "epoch": 4.56,
-      "learning_rate": 6.81556140596846e-05,
-      "loss": 1.9527,
-      "step": 316500
-    },
-    {
-      "epoch": 4.56,
-      "learning_rate": 6.810379117141536e-05,
-      "loss": 1.9581,
-      "step": 317000
-    },
-    {
-      "epoch": 4.57,
-      "learning_rate": 6.805186442966263e-05,
-      "loss": 1.9575,
-      "step": 317500
-    },
-    {
-      "epoch": 4.58,
-      "learning_rate": 6.800014539487691e-05,
-      "loss": 1.9544,
-      "step": 318000
-    },
-    {
-      "epoch": 4.58,
-      "learning_rate": 6.794821865312418e-05,
-      "loss": 1.9547,
-      "step": 318500
-    },
-    {
-      "epoch": 4.59,
-      "learning_rate": 6.789629191137145e-05,
-      "loss": 1.9543,
-      "step": 319000
-    },
-    {
-      "epoch": 4.6,
-      "learning_rate": 6.78443651696187e-05,
-      "loss": 1.9539,
-      "step": 319500
-    },
-    {
-      "epoch": 4.61,
-      "learning_rate": 6.779243842786597e-05,
-      "loss": 1.9576,
-      "step": 320000
-    },
-    {
-      "epoch": 4.61,
-      "learning_rate": 6.774051168611324e-05,
-      "loss": 1.9539,
-      "step": 320500
-    },
-    {
-      "epoch": 4.62,
-      "learning_rate": 6.76885849443605e-05,
-      "loss": 1.9521,
-      "step": 321000
-    },
-    {
-      "epoch": 4.63,
-      "learning_rate": 6.763665820260776e-05,
-      "loss": 1.9576,
-      "step": 321500
-    },
-    {
-      "epoch": 4.63,
-      "learning_rate": 6.758473146085503e-05,
-      "loss": 1.9526,
-      "step": 322000
-    },
-    {
-      "epoch": 4.64,
-      "learning_rate": 6.75328047191023e-05,
-      "loss": 1.9553,
-      "step": 322500
-    },
-    {
-      "epoch": 4.65,
-      "learning_rate": 6.748087797734955e-05,
-      "loss": 1.9532,
-      "step": 323000
-    },
-    {
-      "epoch": 4.66,
-      "learning_rate": 6.742895123559682e-05,
-      "loss": 1.9524,
-      "step": 323500
-    },
-    {
-      "epoch": 4.66,
-      "learning_rate": 6.737712834732759e-05,
-      "loss": 1.9517,
-      "step": 324000
-    },
-    {
-      "epoch": 4.67,
-      "learning_rate": 6.732520160557486e-05,
-      "loss": 1.9524,
-      "step": 324500
-    },
-    {
-      "epoch": 4.68,
-      "learning_rate": 6.727327486382212e-05,
-      "loss": 1.9539,
-      "step": 325000
-    },
-    {
-      "epoch": 4.69,
-      "learning_rate": 6.722134812206938e-05,
-      "loss": 1.9542,
-      "step": 325500
-    },
-    {
-      "epoch": 4.69,
-      "learning_rate": 6.716952523380016e-05,
-      "loss": 1.9531,
-      "step": 326000
-    },
-    {
-      "epoch": 4.7,
-      "learning_rate": 6.711759849204741e-05,
-      "loss": 1.9495,
-      "step": 326500
-    },
-    {
-      "epoch": 4.71,
-      "learning_rate": 6.70656717502947e-05,
-      "loss": 1.9507,
-      "step": 327000
-    },
-    {
-      "epoch": 4.71,
-      "learning_rate": 6.701374500854195e-05,
-      "loss": 1.9461,
-      "step": 327500
-    },
-    {
-      "epoch": 4.72,
-      "learning_rate": 6.696192212027273e-05,
-      "loss": 1.9534,
-      "step": 328000
-    },
-    {
-      "epoch": 4.73,
-      "learning_rate": 6.690999537851998e-05,
-      "loss": 1.9473,
-      "step": 328500
-    },
-    {
-      "epoch": 4.74,
-      "learning_rate": 6.685806863676725e-05,
-      "loss": 1.9534,
-      "step": 329000
-    },
-    {
-      "epoch": 4.74,
-      "learning_rate": 6.680614189501452e-05,
-      "loss": 1.9516,
-      "step": 329500
-    },
-    {
-      "epoch": 4.75,
-      "learning_rate": 6.675431900674529e-05,
-      "loss": 1.9532,
-      "step": 330000
-    },
-    {
-      "epoch": 4.76,
-      "learning_rate": 6.670239226499256e-05,
-      "loss": 1.9505,
-      "step": 330500
-    },
-    {
-      "epoch": 4.76,
-      "learning_rate": 6.665046552323981e-05,
-      "loss": 1.9469,
-      "step": 331000
-    },
-    {
-      "epoch": 4.77,
-      "learning_rate": 6.659853878148709e-05,
-      "loss": 1.9511,
-      "step": 331500
-    },
-    {
-      "epoch": 4.78,
-      "learning_rate": 6.654671589321784e-05,
-      "loss": 1.9506,
-      "step": 332000
-    },
-    {
-      "epoch": 4.79,
-      "learning_rate": 6.649489300494862e-05,
-      "loss": 1.9518,
-      "step": 332500
-    },
-    {
-      "epoch": 4.79,
-      "learning_rate": 6.644296626319588e-05,
-      "loss": 1.9517,
-      "step": 333000
-    },
-    {
-      "epoch": 4.8,
-      "learning_rate": 6.639103952144316e-05,
-      "loss": 1.947,
-      "step": 333500
-    },
-    {
-      "epoch": 4.81,
-      "learning_rate": 6.633911277969042e-05,
-      "loss": 1.9462,
-      "step": 334000
-    },
-    {
-      "epoch": 4.81,
-      "learning_rate": 6.628718603793767e-05,
-      "loss": 1.946,
-      "step": 334500
-    },
-    {
-      "epoch": 4.82,
-      "learning_rate": 6.623525929618495e-05,
-      "loss": 1.9531,
-      "step": 335000
-    },
-    {
-      "epoch": 4.83,
-      "learning_rate": 6.61833325544322e-05,
-      "loss": 1.9506,
-      "step": 335500
-    },
-    {
-      "epoch": 4.84,
-      "learning_rate": 6.613140581267948e-05,
-      "loss": 1.9532,
-      "step": 336000
-    },
-    {
-      "epoch": 4.84,
-      "learning_rate": 6.607958292441024e-05,
-      "loss": 1.9443,
-      "step": 336500
-    },
-    {
-      "epoch": 4.85,
-      "learning_rate": 6.602765618265751e-05,
-      "loss": 1.9486,
-      "step": 337000
-    },
-    {
-      "epoch": 4.86,
-      "learning_rate": 6.597572944090478e-05,
-      "loss": 1.9467,
-      "step": 337500
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 6.592380269915203e-05,
-      "loss": 1.9443,
-      "step": 338000
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 6.587197981088281e-05,
-      "loss": 1.9463,
-      "step": 338500
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 6.582005306913007e-05,
-      "loss": 1.9462,
-      "step": 339000
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.576812632737735e-05,
-      "loss": 1.9507,
-      "step": 339500
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 6.57161995856246e-05,
-      "loss": 1.9489,
-      "step": 340000
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 6.566448055083888e-05,
-      "loss": 1.9447,
-      "step": 340500
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 6.561255380908614e-05,
-      "loss": 1.9469,
-      "step": 341000
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 6.556062706733342e-05,
-      "loss": 1.9456,
-      "step": 341500
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 6.550870032558067e-05,
-      "loss": 1.9441,
-      "step": 342000
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 6.545677358382794e-05,
-      "loss": 1.9454,
-      "step": 342500
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 6.540484684207521e-05,
-      "loss": 1.947,
-      "step": 343000
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 6.535292010032246e-05,
-      "loss": 1.9427,
-      "step": 343500
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 6.530099335856973e-05,
-      "loss": 1.9461,
-      "step": 344000
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 6.52491704703005e-05,
-      "loss": 1.9465,
-      "step": 344500
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.519724372854777e-05,
-      "loss": 1.9443,
-      "step": 345000
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.514531698679504e-05,
-      "loss": 1.9455,
-      "step": 345500
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 6.509339024504229e-05,
-      "loss": 1.9419,
-      "step": 346000
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 6.504156735677307e-05,
-      "loss": 1.9396,
-      "step": 346500
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 6.498964061502032e-05,
-      "loss": 1.937,
-      "step": 347000
-    },
-    {
-      "epoch": 5.0,
-      "eval_accuracy": 0.6320096391256435,
-      "eval_loss": 1.807983636856079,
-      "eval_runtime": 648.1194,
-      "eval_samples_per_second": 831.54,
-      "eval_steps_per_second": 34.648,
-      "step": 347365
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 6.49377138732676e-05,
-      "loss": 1.9433,
-      "step": 347500
-    },
-    {
-      "epoch": 5.01,
-      "learning_rate": 6.488589098499836e-05,
-      "loss": 1.9431,
-      "step": 348000
-    },
-    {
-      "epoch": 5.02,
-      "learning_rate": 6.483396424324564e-05,
-      "loss": 1.9447,
-      "step": 348500
-    },
-    {
-      "epoch": 5.02,
-      "learning_rate": 6.47820375014929e-05,
-      "loss": 1.9422,
-      "step": 349000
-    },
-    {
-      "epoch": 5.03,
-      "learning_rate": 6.473011075974016e-05,
-      "loss": 1.9421,
-      "step": 349500
-    },
-    {
-      "epoch": 5.04,
-      "learning_rate": 6.467818401798743e-05,
-      "loss": 1.94,
-      "step": 350000
-    },
-    {
-      "epoch": 5.05,
-      "learning_rate": 6.462625727623469e-05,
-      "loss": 1.9386,
-      "step": 350500
-    },
-    {
-      "epoch": 5.05,
-      "learning_rate": 6.457433053448196e-05,
-      "loss": 1.9419,
-      "step": 351000
-    },
-    {
-      "epoch": 5.06,
-      "learning_rate": 6.452250764621272e-05,
-      "loss": 1.9373,
-      "step": 351500
-    },
-    {
-      "epoch": 5.07,
-      "learning_rate": 6.447058090445999e-05,
-      "loss": 1.9377,
-      "step": 352000
-    },
-    {
-      "epoch": 5.07,
-      "learning_rate": 6.441865416270726e-05,
-      "loss": 1.9414,
-      "step": 352500
-    },
-    {
-      "epoch": 5.08,
-      "learning_rate": 6.436672742095453e-05,
-      "loss": 1.9413,
-      "step": 353000
-    },
-    {
-      "epoch": 5.09,
-      "learning_rate": 6.431480067920178e-05,
-      "loss": 1.9364,
-      "step": 353500
-    },
-    {
-      "epoch": 5.1,
-      "learning_rate": 6.426287393744905e-05,
-      "loss": 1.9375,
-      "step": 354000
-    },
-    {
-      "epoch": 5.1,
-      "learning_rate": 6.421094719569632e-05,
-      "loss": 1.9397,
-      "step": 354500
-    },
-    {
-      "epoch": 5.11,
-      "learning_rate": 6.415902045394357e-05,
-      "loss": 1.9358,
-      "step": 355000
-    },
-    {
-      "epoch": 5.12,
-      "learning_rate": 6.410719756567435e-05,
-      "loss": 1.9371,
-      "step": 355500
-    },
-    {
-      "epoch": 5.12,
-      "learning_rate": 6.405527082392161e-05,
-      "loss": 1.9376,
-      "step": 356000
-    },
-    {
-      "epoch": 5.13,
-      "learning_rate": 6.400334408216888e-05,
-      "loss": 1.9338,
-      "step": 356500
-    },
-    {
-      "epoch": 5.14,
-      "learning_rate": 6.395141734041614e-05,
-      "loss": 1.9394,
-      "step": 357000
-    },
-    {
-      "epoch": 5.15,
-      "learning_rate": 6.389959445214691e-05,
-      "loss": 1.9397,
-      "step": 357500
-    },
-    {
-      "epoch": 5.15,
-      "learning_rate": 6.384766771039418e-05,
-      "loss": 1.9415,
-      "step": 358000
-    },
-    {
-      "epoch": 5.16,
-      "learning_rate": 6.379574096864145e-05,
-      "loss": 1.9385,
-      "step": 358500
-    },
-    {
-      "epoch": 5.17,
-      "learning_rate": 6.374381422688871e-05,
-      "loss": 1.9421,
-      "step": 359000
-    },
-    {
-      "epoch": 5.17,
-      "learning_rate": 6.369199133861948e-05,
-      "loss": 1.9324,
-      "step": 359500
-    },
-    {
-      "epoch": 5.18,
-      "learning_rate": 6.364006459686675e-05,
-      "loss": 1.9355,
-      "step": 360000
-    },
-    {
-      "epoch": 5.19,
-      "learning_rate": 6.3588137855114e-05,
-      "loss": 1.9382,
-      "step": 360500
-    },
-    {
-      "epoch": 5.2,
-      "learning_rate": 6.353621111336127e-05,
-      "loss": 1.9357,
-      "step": 361000
-    },
-    {
-      "epoch": 5.2,
-      "learning_rate": 6.348438822509204e-05,
-      "loss": 1.9368,
-      "step": 361500
-    },
-    {
-      "epoch": 5.21,
-      "learning_rate": 6.34324614833393e-05,
-      "loss": 1.9403,
-      "step": 362000
-    },
-    {
-      "epoch": 5.22,
-      "learning_rate": 6.338053474158658e-05,
-      "loss": 1.9365,
-      "step": 362500
-    },
-    {
-      "epoch": 5.23,
-      "learning_rate": 6.332860799983383e-05,
-      "loss": 1.936,
-      "step": 363000
-    },
-    {
-      "epoch": 5.23,
-      "learning_rate": 6.327678511156461e-05,
-      "loss": 1.9344,
-      "step": 363500
-    },
-    {
-      "epoch": 5.24,
-      "learning_rate": 6.322485836981186e-05,
-      "loss": 1.9368,
-      "step": 364000
-    },
-    {
-      "epoch": 5.25,
-      "learning_rate": 6.317293162805915e-05,
-      "loss": 1.9329,
-      "step": 364500
-    },
-    {
-      "epoch": 5.25,
-      "learning_rate": 6.31210048863064e-05,
-      "loss": 1.9337,
-      "step": 365000
-    },
-    {
-      "epoch": 5.26,
-      "learning_rate": 6.306918199803718e-05,
-      "loss": 1.9338,
-      "step": 365500
-    },
-    {
-      "epoch": 5.27,
-      "learning_rate": 6.301725525628444e-05,
-      "loss": 1.9318,
-      "step": 366000
-    },
-    {
-      "epoch": 5.28,
-      "learning_rate": 6.296532851453169e-05,
-      "loss": 1.9368,
-      "step": 366500
-    },
-    {
-      "epoch": 5.28,
-      "learning_rate": 6.291340177277897e-05,
-      "loss": 1.9319,
-      "step": 367000
-    },
-    {
-      "epoch": 5.29,
-      "learning_rate": 6.286168273799325e-05,
-      "loss": 1.9398,
-      "step": 367500
-    },
-    {
-      "epoch": 5.3,
-      "learning_rate": 6.28097559962405e-05,
-      "loss": 1.9405,
-      "step": 368000
-    },
-    {
-      "epoch": 5.3,
-      "learning_rate": 6.275782925448777e-05,
-      "loss": 1.9373,
-      "step": 368500
-    },
-    {
-      "epoch": 5.31,
-      "learning_rate": 6.270590251273504e-05,
-      "loss": 1.9335,
-      "step": 369000
-    },
-    {
-      "epoch": 5.32,
-      "learning_rate": 6.26539757709823e-05,
-      "loss": 1.937,
-      "step": 369500
-    },
-    {
-      "epoch": 5.33,
-      "learning_rate": 6.260204902922956e-05,
-      "loss": 1.9337,
-      "step": 370000
-    },
-    {
-      "epoch": 5.33,
-      "learning_rate": 6.255012228747683e-05,
-      "loss": 1.9331,
-      "step": 370500
-    },
-    {
-      "epoch": 5.34,
-      "learning_rate": 6.249819554572409e-05,
-      "loss": 1.9328,
-      "step": 371000
-    },
-    {
-      "epoch": 5.35,
-      "learning_rate": 6.244637265745487e-05,
-      "loss": 1.9377,
-      "step": 371500
-    },
-    {
-      "epoch": 5.35,
-      "learning_rate": 6.239444591570212e-05,
-      "loss": 1.9352,
-      "step": 372000
-    },
-    {
-      "epoch": 5.36,
-      "learning_rate": 6.23425191739494e-05,
-      "loss": 1.9263,
-      "step": 372500
-    },
-    {
-      "epoch": 5.37,
-      "learning_rate": 6.229059243219666e-05,
-      "loss": 1.9342,
-      "step": 373000
-    },
-    {
-      "epoch": 5.38,
-      "learning_rate": 6.223876954392744e-05,
-      "loss": 1.9368,
-      "step": 373500
-    },
-    {
-      "epoch": 5.38,
-      "learning_rate": 6.218684280217469e-05,
-      "loss": 1.9308,
-      "step": 374000
-    },
-    {
-      "epoch": 5.39,
-      "learning_rate": 6.213491606042196e-05,
-      "loss": 1.93,
-      "step": 374500
-    },
-    {
-      "epoch": 5.4,
-      "learning_rate": 6.208298931866923e-05,
-      "loss": 1.9335,
-      "step": 375000
-    },
-    {
-      "epoch": 5.4,
-      "learning_rate": 6.20311664304e-05,
-      "loss": 1.933,
-      "step": 375500
-    },
-    {
-      "epoch": 5.41,
-      "learning_rate": 6.197923968864726e-05,
-      "loss": 1.9321,
-      "step": 376000
-    },
-    {
-      "epoch": 5.42,
-      "learning_rate": 6.192731294689452e-05,
-      "loss": 1.9297,
-      "step": 376500
-    },
-    {
-      "epoch": 5.43,
-      "learning_rate": 6.18753862051418e-05,
-      "loss": 1.9342,
-      "step": 377000
-    },
-    {
-      "epoch": 5.43,
-      "learning_rate": 6.182356331687255e-05,
-      "loss": 1.931,
-      "step": 377500
-    },
-    {
-      "epoch": 5.44,
-      "learning_rate": 6.177163657511983e-05,
-      "loss": 1.932,
-      "step": 378000
-    },
-    {
-      "epoch": 5.45,
-      "learning_rate": 6.171970983336709e-05,
-      "loss": 1.9308,
-      "step": 378500
-    },
-    {
-      "epoch": 5.46,
-      "learning_rate": 6.166778309161434e-05,
-      "loss": 1.9328,
-      "step": 379000
-    },
-    {
-      "epoch": 5.46,
-      "learning_rate": 6.161596020334512e-05,
-      "loss": 1.9282,
-      "step": 379500
-    },
-    {
-      "epoch": 5.47,
-      "learning_rate": 6.156403346159238e-05,
-      "loss": 1.9306,
-      "step": 380000
-    },
-    {
-      "epoch": 5.48,
-      "learning_rate": 6.151210671983966e-05,
-      "loss": 1.9305,
-      "step": 380500
-    },
-    {
-      "epoch": 5.48,
-      "learning_rate": 6.146017997808692e-05,
-      "loss": 1.9338,
-      "step": 381000
-    },
-    {
-      "epoch": 5.49,
-      "learning_rate": 6.14083570898177e-05,
-      "loss": 1.9293,
-      "step": 381500
-    },
-    {
-      "epoch": 5.5,
-      "learning_rate": 6.135643034806495e-05,
-      "loss": 1.9292,
-      "step": 382000
-    },
-    {
-      "epoch": 5.51,
-      "learning_rate": 6.130450360631222e-05,
-      "loss": 1.9307,
-      "step": 382500
-    },
-    {
-      "epoch": 5.51,
-      "learning_rate": 6.125257686455949e-05,
-      "loss": 1.9307,
-      "step": 383000
-    },
-    {
-      "epoch": 5.52,
-      "learning_rate": 6.120075397629025e-05,
-      "loss": 1.9323,
-      "step": 383500
-    },
-    {
-      "epoch": 5.53,
-      "learning_rate": 6.114882723453752e-05,
-      "loss": 1.9321,
-      "step": 384000
-    },
-    {
-      "epoch": 5.53,
-      "learning_rate": 6.109690049278478e-05,
-      "loss": 1.9339,
-      "step": 384500
-    },
-    {
-      "epoch": 5.54,
-      "learning_rate": 6.104497375103204e-05,
-      "loss": 1.9289,
-      "step": 385000
-    },
-    {
-      "epoch": 5.55,
-      "learning_rate": 6.099315086276282e-05,
-      "loss": 1.9275,
-      "step": 385500
-    },
-    {
-      "epoch": 5.56,
-      "learning_rate": 6.094122412101008e-05,
-      "loss": 1.9243,
-      "step": 386000
-    },
-    {
-      "epoch": 5.56,
-      "learning_rate": 6.088929737925735e-05,
-      "loss": 1.9277,
-      "step": 386500
-    },
-    {
-      "epoch": 5.57,
-      "learning_rate": 6.083737063750461e-05,
-      "loss": 1.9299,
-      "step": 387000
-    },
-    {
-      "epoch": 5.58,
-      "learning_rate": 6.0785443895751884e-05,
-      "loss": 1.9301,
-      "step": 387500
-    },
-    {
-      "epoch": 5.58,
-      "learning_rate": 6.073362100748264e-05,
-      "loss": 1.9271,
-      "step": 388000
-    },
-    {
-      "epoch": 5.59,
-      "learning_rate": 6.068169426572992e-05,
-      "loss": 1.9295,
-      "step": 388500
-    },
-    {
-      "epoch": 5.6,
-      "learning_rate": 6.062976752397718e-05,
-      "loss": 1.9269,
-      "step": 389000
-    },
-    {
-      "epoch": 5.61,
-      "learning_rate": 6.0577840782224434e-05,
-      "loss": 1.9251,
-      "step": 389500
-    },
-    {
-      "epoch": 5.61,
-      "learning_rate": 6.052591404047171e-05,
-      "loss": 1.9253,
-      "step": 390000
-    },
-    {
-      "epoch": 5.62,
-      "learning_rate": 6.047398729871897e-05,
-      "loss": 1.9297,
-      "step": 390500
-    },
-    {
-      "epoch": 5.63,
-      "learning_rate": 6.042206055696623e-05,
-      "loss": 1.9286,
-      "step": 391000
-    },
-    {
-      "epoch": 5.64,
-      "learning_rate": 6.03701338152135e-05,
-      "loss": 1.9254,
-      "step": 391500
-    },
-    {
-      "epoch": 5.64,
-      "learning_rate": 6.031820707346076e-05,
-      "loss": 1.9244,
-      "step": 392000
-    },
-    {
-      "epoch": 5.65,
-      "learning_rate": 6.0266384185191535e-05,
-      "loss": 1.9291,
-      "step": 392500
-    },
-    {
-      "epoch": 5.66,
-      "learning_rate": 6.02144574434388e-05,
-      "loss": 1.925,
-      "step": 393000
-    },
-    {
-      "epoch": 5.66,
-      "learning_rate": 6.016253070168606e-05,
-      "loss": 1.924,
-      "step": 393500
-    },
-    {
-      "epoch": 5.67,
-      "learning_rate": 6.0110603959933333e-05,
-      "loss": 1.9275,
-      "step": 394000
-    },
-    {
-      "epoch": 5.68,
-      "learning_rate": 6.005878107166409e-05,
-      "loss": 1.9278,
-      "step": 394500
-    },
-    {
-      "epoch": 5.69,
-      "learning_rate": 6.0006958183394866e-05,
-      "loss": 1.9219,
-      "step": 395000
-    },
-    {
-      "epoch": 5.69,
-      "learning_rate": 5.995503144164213e-05,
-      "loss": 1.9268,
-      "step": 395500
-    },
-    {
-      "epoch": 5.7,
-      "learning_rate": 5.99031046998894e-05,
-      "loss": 1.9273,
-      "step": 396000
-    },
-    {
-      "epoch": 5.71,
-      "learning_rate": 5.9851177958136664e-05,
-      "loss": 1.9229,
-      "step": 396500
-    },
-    {
-      "epoch": 5.71,
-      "learning_rate": 5.979925121638393e-05,
-      "loss": 1.9265,
-      "step": 397000
-    },
-    {
-      "epoch": 5.72,
-      "learning_rate": 5.9747324474631194e-05,
-      "loss": 1.9246,
-      "step": 397500
-    },
-    {
-      "epoch": 5.73,
-      "learning_rate": 5.9695397732878455e-05,
-      "loss": 1.9264,
-      "step": 398000
-    },
-    {
-      "epoch": 5.74,
-      "learning_rate": 5.964347099112573e-05,
-      "loss": 1.9238,
-      "step": 398500
-    },
-    {
-      "epoch": 5.74,
-      "learning_rate": 5.959164810285649e-05,
-      "loss": 1.9216,
-      "step": 399000
-    },
-    {
-      "epoch": 5.75,
-      "learning_rate": 5.9539721361103765e-05,
-      "loss": 1.9214,
-      "step": 399500
-    },
-    {
-      "epoch": 5.76,
-      "learning_rate": 5.948779461935102e-05,
-      "loss": 1.9282,
-      "step": 400000
-    },
-    {
-      "epoch": 5.76,
-      "learning_rate": 5.943586787759828e-05,
-      "loss": 1.9277,
-      "step": 400500
-    },
-    {
-      "epoch": 5.77,
-      "learning_rate": 5.9384044989329054e-05,
-      "loss": 1.9213,
-      "step": 401000
-    },
-    {
-      "epoch": 5.78,
-      "learning_rate": 5.9332118247576316e-05,
-      "loss": 1.924,
-      "step": 401500
-    },
-    {
-      "epoch": 5.79,
-      "learning_rate": 5.928019150582359e-05,
-      "loss": 1.9252,
-      "step": 402000
-    },
-    {
-      "epoch": 5.79,
-      "learning_rate": 5.922826476407085e-05,
-      "loss": 1.925,
-      "step": 402500
-    },
-    {
-      "epoch": 5.8,
-      "learning_rate": 5.9176441875801625e-05,
-      "loss": 1.9211,
-      "step": 403000
-    },
-    {
-      "epoch": 5.81,
-      "learning_rate": 5.912451513404889e-05,
-      "loss": 1.9232,
-      "step": 403500
-    },
-    {
-      "epoch": 5.82,
-      "learning_rate": 5.907258839229615e-05,
-      "loss": 1.9179,
-      "step": 404000
-    },
-    {
-      "epoch": 5.82,
-      "learning_rate": 5.902066165054342e-05,
-      "loss": 1.9199,
-      "step": 404500
-    },
-    {
-      "epoch": 5.83,
-      "learning_rate": 5.896883876227418e-05,
-      "loss": 1.9233,
-      "step": 405000
-    },
-    {
-      "epoch": 5.84,
-      "learning_rate": 5.891691202052145e-05,
-      "loss": 1.9235,
-      "step": 405500
-    },
-    {
-      "epoch": 5.84,
-      "learning_rate": 5.886498527876871e-05,
-      "loss": 1.9218,
-      "step": 406000
-    },
-    {
-      "epoch": 5.85,
-      "learning_rate": 5.881305853701599e-05,
-      "loss": 1.9206,
-      "step": 406500
-    },
-    {
-      "epoch": 5.86,
-      "learning_rate": 5.876123564874675e-05,
-      "loss": 1.9183,
-      "step": 407000
-    },
-    {
-      "epoch": 5.87,
-      "learning_rate": 5.870930890699402e-05,
-      "loss": 1.9202,
-      "step": 407500
-    },
-    {
-      "epoch": 5.87,
-      "learning_rate": 5.8657382165241284e-05,
-      "loss": 1.9221,
-      "step": 408000
-    },
-    {
-      "epoch": 5.88,
-      "learning_rate": 5.8605455423488545e-05,
-      "loss": 1.9206,
-      "step": 408500
-    },
-    {
-      "epoch": 5.89,
-      "learning_rate": 5.855363253521932e-05,
-      "loss": 1.9222,
-      "step": 409000
-    },
-    {
-      "epoch": 5.89,
-      "learning_rate": 5.850170579346658e-05,
-      "loss": 1.9192,
-      "step": 409500
-    },
-    {
-      "epoch": 5.9,
-      "learning_rate": 5.844977905171385e-05,
-      "loss": 1.9204,
-      "step": 410000
-    },
-    {
-      "epoch": 5.91,
-      "learning_rate": 5.839785230996111e-05,
-      "loss": 1.9208,
-      "step": 410500
-    },
-    {
-      "epoch": 5.92,
-      "learning_rate": 5.834602942169188e-05,
-      "loss": 1.9199,
-      "step": 411000
-    },
-    {
-      "epoch": 5.92,
-      "learning_rate": 5.8294102679939144e-05,
-      "loss": 1.9192,
-      "step": 411500
-    },
-    {
-      "epoch": 5.93,
-      "learning_rate": 5.8242175938186406e-05,
-      "loss": 1.9218,
-      "step": 412000
-    },
-    {
-      "epoch": 5.94,
-      "learning_rate": 5.8190249196433674e-05,
-      "loss": 1.9216,
-      "step": 412500
-    },
-    {
-      "epoch": 5.94,
-      "learning_rate": 5.813842630816444e-05,
-      "loss": 1.9223,
-      "step": 413000
-    },
-    {
-      "epoch": 5.95,
-      "learning_rate": 5.808660341989521e-05,
-      "loss": 1.9185,
-      "step": 413500
-    },
-    {
-      "epoch": 5.96,
-      "learning_rate": 5.8034676678142475e-05,
-      "loss": 1.925,
-      "step": 414000
-    },
-    {
-      "epoch": 5.97,
-      "learning_rate": 5.798274993638974e-05,
-      "loss": 1.9156,
-      "step": 414500
-    },
-    {
-      "epoch": 5.97,
-      "learning_rate": 5.7930823194637005e-05,
-      "loss": 1.9226,
-      "step": 415000
-    },
-    {
-      "epoch": 5.98,
-      "learning_rate": 5.7878896452884266e-05,
-      "loss": 1.9174,
-      "step": 415500
-    },
-    {
-      "epoch": 5.99,
-      "learning_rate": 5.782696971113154e-05,
-      "loss": 1.9211,
-      "step": 416000
-    },
-    {
-      "epoch": 6.0,
-      "learning_rate": 5.77750429693788e-05,
-      "loss": 1.9152,
-      "step": 416500
-    },
-    {
-      "epoch": 6.0,
-      "eval_accuracy": 0.6361376420590098,
-      "eval_loss": 1.781764030456543,
-      "eval_runtime": 645.9143,
-      "eval_samples_per_second": 834.378,
-      "eval_steps_per_second": 34.766,
-      "step": 416838
-    },
-    {
-      "epoch": 6.0,
-      "learning_rate": 5.772311622762607e-05,
-      "loss": 1.9179,
-      "step": 417000
-    },
-    {
-      "epoch": 6.01,
-      "learning_rate": 5.767129333935684e-05,
-      "loss": 1.9186,
-      "step": 417500
-    },
-    {
-      "epoch": 6.02,
-      "learning_rate": 5.761947045108761e-05,
-      "loss": 1.9127,
-      "step": 418000
-    },
-    {
-      "epoch": 6.02,
-      "learning_rate": 5.756754370933487e-05,
-      "loss": 1.9154,
-      "step": 418500
-    },
-    {
-      "epoch": 6.03,
-      "learning_rate": 5.751561696758214e-05,
-      "loss": 1.9157,
-      "step": 419000
-    },
-    {
-      "epoch": 6.04,
-      "learning_rate": 5.74636902258294e-05,
-      "loss": 1.912,
-      "step": 419500
-    },
-    {
-      "epoch": 6.05,
-      "learning_rate": 5.741176348407666e-05,
-      "loss": 1.9186,
-      "step": 420000
-    },
-    {
-      "epoch": 6.05,
-      "learning_rate": 5.735983674232394e-05,
-      "loss": 1.9151,
-      "step": 420500
-    },
-    {
-      "epoch": 6.06,
-      "learning_rate": 5.73079100005712e-05,
-      "loss": 1.9187,
-      "step": 421000
-    },
-    {
-      "epoch": 6.07,
-      "learning_rate": 5.7255983258818454e-05,
-      "loss": 1.913,
-      "step": 421500
-    },
-    {
-      "epoch": 6.07,
-      "learning_rate": 5.7204160370549234e-05,
-      "loss": 1.9113,
-      "step": 422000
-    },
-    {
-      "epoch": 6.08,
-      "learning_rate": 5.715223362879649e-05,
-      "loss": 1.9153,
-      "step": 422500
-    },
-    {
-      "epoch": 6.09,
-      "learning_rate": 5.7100306887043764e-05,
-      "loss": 1.9138,
-      "step": 423000
-    },
-    {
-      "epoch": 6.1,
-      "learning_rate": 5.7048380145291026e-05,
-      "loss": 1.9142,
-      "step": 423500
-    },
-    {
-      "epoch": 6.1,
-      "learning_rate": 5.69965572570218e-05,
-      "loss": 1.9134,
-      "step": 424000
-    },
-    {
-      "epoch": 6.11,
-      "learning_rate": 5.694463051526906e-05,
-      "loss": 1.9176,
-      "step": 424500
-    },
-    {
-      "epoch": 6.12,
-      "learning_rate": 5.689270377351632e-05,
-      "loss": 1.9119,
-      "step": 425000
-    },
-    {
-      "epoch": 6.12,
-      "learning_rate": 5.684077703176359e-05,
-      "loss": 1.918,
-      "step": 425500
-    },
-    {
-      "epoch": 6.13,
-      "learning_rate": 5.678895414349436e-05,
-      "loss": 1.9135,
-      "step": 426000
-    },
-    {
-      "epoch": 6.14,
-      "learning_rate": 5.673713125522513e-05,
-      "loss": 1.9183,
-      "step": 426500
-    },
-    {
-      "epoch": 6.15,
-      "learning_rate": 5.66852045134724e-05,
-      "loss": 1.9146,
-      "step": 427000
-    },
-    {
-      "epoch": 6.15,
-      "learning_rate": 5.663327777171966e-05,
-      "loss": 1.9156,
-      "step": 427500
-    },
-    {
-      "epoch": 6.16,
-      "learning_rate": 5.658135102996692e-05,
-      "loss": 1.917,
-      "step": 428000
-    },
-    {
-      "epoch": 6.17,
-      "learning_rate": 5.6529424288214196e-05,
-      "loss": 1.914,
-      "step": 428500
-    },
-    {
-      "epoch": 6.18,
-      "learning_rate": 5.647749754646146e-05,
-      "loss": 1.9172,
-      "step": 429000
-    },
-    {
-      "epoch": 6.18,
-      "learning_rate": 5.642557080470872e-05,
-      "loss": 1.9101,
-      "step": 429500
-    },
-    {
-      "epoch": 6.19,
-      "learning_rate": 5.637364406295599e-05,
-      "loss": 1.9107,
-      "step": 430000
-    },
-    {
-      "epoch": 6.2,
-      "learning_rate": 5.632182117468675e-05,
-      "loss": 1.9096,
-      "step": 430500
-    },
-    {
-      "epoch": 6.2,
-      "learning_rate": 5.626989443293402e-05,
-      "loss": 1.9118,
-      "step": 431000
-    },
-    {
-      "epoch": 6.21,
-      "learning_rate": 5.621796769118128e-05,
-      "loss": 1.9104,
-      "step": 431500
-    },
-    {
-      "epoch": 6.22,
-      "learning_rate": 5.6166040949428544e-05,
-      "loss": 1.9105,
-      "step": 432000
-    },
-    {
-      "epoch": 6.23,
-      "learning_rate": 5.611421806115932e-05,
-      "loss": 1.9169,
-      "step": 432500
-    },
-    {
-      "epoch": 6.23,
-      "learning_rate": 5.606229131940658e-05,
-      "loss": 1.914,
-      "step": 433000
-    },
-    {
-      "epoch": 6.24,
-      "learning_rate": 5.6010364577653854e-05,
-      "loss": 1.9136,
-      "step": 433500
-    },
-    {
-      "epoch": 6.25,
-      "learning_rate": 5.595843783590111e-05,
-      "loss": 1.9097,
-      "step": 434000
-    },
-    {
-      "epoch": 6.25,
-      "learning_rate": 5.590661494763189e-05,
-      "loss": 1.9132,
-      "step": 434500
-    },
-    {
-      "epoch": 6.26,
-      "learning_rate": 5.585468820587914e-05,
-      "loss": 1.9113,
-      "step": 435000
-    },
-    {
-      "epoch": 6.27,
-      "learning_rate": 5.580276146412642e-05,
-      "loss": 1.9137,
-      "step": 435500
-    },
-    {
-      "epoch": 6.28,
-      "learning_rate": 5.575083472237368e-05,
-      "loss": 1.9082,
-      "step": 436000
-    },
-    {
-      "epoch": 6.28,
-      "learning_rate": 5.569901183410445e-05,
-      "loss": 1.9129,
-      "step": 436500
-    },
-    {
-      "epoch": 6.29,
-      "learning_rate": 5.5647085092351714e-05,
-      "loss": 1.9088,
-      "step": 437000
-    },
-    {
-      "epoch": 6.3,
-      "learning_rate": 5.5595158350598976e-05,
-      "loss": 1.9119,
-      "step": 437500
-    },
-    {
-      "epoch": 6.3,
-      "learning_rate": 5.5543231608846244e-05,
-      "loss": 1.9125,
-      "step": 438000
-    },
-    {
-      "epoch": 6.31,
-      "learning_rate": 5.549140872057701e-05,
-      "loss": 1.909,
-      "step": 438500
-    },
-    {
-      "epoch": 6.32,
-      "learning_rate": 5.543948197882428e-05,
-      "loss": 1.9111,
-      "step": 439000
-    },
-    {
-      "epoch": 6.33,
-      "learning_rate": 5.538755523707154e-05,
-      "loss": 1.9112,
-      "step": 439500
-    },
-    {
-      "epoch": 6.33,
-      "learning_rate": 5.53356284953188e-05,
-      "loss": 1.9093,
-      "step": 440000
-    },
-    {
-      "epoch": 6.34,
-      "learning_rate": 5.5283805607049575e-05,
-      "loss": 1.9096,
-      "step": 440500
-    },
-    {
-      "epoch": 6.35,
-      "learning_rate": 5.5231878865296836e-05,
-      "loss": 1.9128,
-      "step": 441000
-    },
-    {
-      "epoch": 6.35,
-      "learning_rate": 5.517995212354411e-05,
-      "loss": 1.9075,
-      "step": 441500
-    },
-    {
-      "epoch": 6.36,
-      "learning_rate": 5.512802538179137e-05,
-      "loss": 1.9103,
-      "step": 442000
-    },
-    {
-      "epoch": 6.37,
-      "learning_rate": 5.5076202493522146e-05,
-      "loss": 1.9123,
-      "step": 442500
-    },
-    {
-      "epoch": 6.38,
-      "learning_rate": 5.502427575176941e-05,
-      "loss": 1.9121,
-      "step": 443000
-    },
-    {
-      "epoch": 6.38,
-      "learning_rate": 5.497245286350018e-05,
-      "loss": 1.9133,
-      "step": 443500
-    },
-    {
-      "epoch": 6.39,
-      "learning_rate": 5.492052612174744e-05,
-      "loss": 1.9112,
-      "step": 444000
-    },
-    {
-      "epoch": 6.4,
-      "learning_rate": 5.4868599379994704e-05,
-      "loss": 1.9081,
-      "step": 444500
-    },
-    {
-      "epoch": 6.41,
-      "learning_rate": 5.481667263824197e-05,
-      "loss": 1.9141,
-      "step": 445000
-    },
-    {
-      "epoch": 6.41,
-      "learning_rate": 5.476474589648923e-05,
-      "loss": 1.9113,
-      "step": 445500
-    },
-    {
-      "epoch": 6.42,
-      "learning_rate": 5.471281915473651e-05,
-      "loss": 1.9081,
-      "step": 446000
-    },
-    {
-      "epoch": 6.43,
-      "learning_rate": 5.466089241298376e-05,
-      "loss": 1.9097,
-      "step": 446500
-    },
-    {
-      "epoch": 6.43,
-      "learning_rate": 5.4608965671231025e-05,
-      "loss": 1.9084,
-      "step": 447000
-    },
-    {
-      "epoch": 6.44,
-      "learning_rate": 5.45571427829618e-05,
-      "loss": 1.9084,
-      "step": 447500
-    },
-    {
-      "epoch": 6.45,
-      "learning_rate": 5.450521604120906e-05,
-      "loss": 1.9099,
-      "step": 448000
-    },
-    {
-      "epoch": 6.46,
-      "learning_rate": 5.4453289299456334e-05,
-      "loss": 1.9089,
-      "step": 448500
-    },
-    {
-      "epoch": 6.46,
-      "learning_rate": 5.4401466411187094e-05,
-      "loss": 1.9058,
-      "step": 449000
-    },
-    {
-      "epoch": 6.47,
-      "learning_rate": 5.434953966943437e-05,
-      "loss": 1.91,
-      "step": 449500
-    },
-    {
-      "epoch": 6.48,
-      "learning_rate": 5.429761292768163e-05,
-      "loss": 1.9086,
-      "step": 450000
-    },
-    {
-      "epoch": 6.48,
-      "learning_rate": 5.424568618592889e-05,
-      "loss": 1.9065,
-      "step": 450500
-    },
-    {
-      "epoch": 6.49,
-      "learning_rate": 5.419375944417616e-05,
-      "loss": 1.9076,
-      "step": 451000
-    },
-    {
-      "epoch": 6.5,
-      "learning_rate": 5.414183270242342e-05,
-      "loss": 1.9092,
-      "step": 451500
-    },
-    {
-      "epoch": 6.51,
-      "learning_rate": 5.408990596067068e-05,
-      "loss": 1.9111,
-      "step": 452000
-    },
-    {
-      "epoch": 6.51,
-      "learning_rate": 5.403797921891796e-05,
-      "loss": 1.9142,
-      "step": 452500
-    },
-    {
-      "epoch": 6.52,
-      "learning_rate": 5.398615633064872e-05,
-      "loss": 1.9037,
-      "step": 453000
-    },
-    {
-      "epoch": 6.53,
-      "learning_rate": 5.393422958889599e-05,
-      "loss": 1.9089,
-      "step": 453500
-    },
-    {
-      "epoch": 6.53,
-      "learning_rate": 5.3882302847143254e-05,
-      "loss": 1.9084,
-      "step": 454000
-    },
-    {
-      "epoch": 6.54,
-      "learning_rate": 5.383037610539052e-05,
-      "loss": 1.9075,
-      "step": 454500
-    },
-    {
-      "epoch": 6.55,
-      "learning_rate": 5.377855321712129e-05,
-      "loss": 1.9075,
-      "step": 455000
-    },
-    {
-      "epoch": 6.56,
-      "learning_rate": 5.372662647536856e-05,
-      "loss": 1.9099,
-      "step": 455500
-    },
-    {
-      "epoch": 6.56,
-      "learning_rate": 5.367480358709932e-05,
-      "loss": 1.9025,
-      "step": 456000
-    },
-    {
-      "epoch": 6.57,
-      "learning_rate": 5.362287684534659e-05,
-      "loss": 1.9094,
-      "step": 456500
-    },
-    {
-      "epoch": 6.58,
-      "learning_rate": 5.357095010359385e-05,
-      "loss": 1.9018,
-      "step": 457000
-    },
-    {
-      "epoch": 6.59,
-      "learning_rate": 5.3519023361841115e-05,
-      "loss": 1.9076,
-      "step": 457500
-    },
-    {
-      "epoch": 6.59,
-      "learning_rate": 5.346709662008838e-05,
-      "loss": 1.9071,
-      "step": 458000
-    },
-    {
-      "epoch": 6.6,
-      "learning_rate": 5.3415169878335644e-05,
-      "loss": 1.9096,
-      "step": 458500
-    },
-    {
-      "epoch": 6.61,
-      "learning_rate": 5.336334699006642e-05,
-      "loss": 1.9021,
-      "step": 459000
-    },
-    {
-      "epoch": 6.61,
-      "learning_rate": 5.331142024831368e-05,
-      "loss": 1.9016,
-      "step": 459500
-    },
-    {
-      "epoch": 6.62,
-      "learning_rate": 5.325949350656094e-05,
-      "loss": 1.9048,
-      "step": 460000
-    },
-    {
-      "epoch": 6.63,
-      "learning_rate": 5.3207566764808216e-05,
-      "loss": 1.9037,
-      "step": 460500
-    },
-    {
-      "epoch": 6.64,
-      "learning_rate": 5.315564002305548e-05,
-      "loss": 1.9029,
-      "step": 461000
-    },
-    {
-      "epoch": 6.64,
-      "learning_rate": 5.310371328130274e-05,
-      "loss": 1.9028,
-      "step": 461500
-    },
-    {
-      "epoch": 6.65,
-      "learning_rate": 5.305178653955001e-05,
-      "loss": 1.9073,
-      "step": 462000
-    },
-    {
-      "epoch": 6.66,
-      "learning_rate": 5.299985979779727e-05,
-      "loss": 1.9047,
-      "step": 462500
-    },
-    {
-      "epoch": 6.66,
-      "learning_rate": 5.294803690952804e-05,
-      "loss": 1.9033,
-      "step": 463000
-    },
-    {
-      "epoch": 6.67,
-      "learning_rate": 5.289621402125881e-05,
-      "loss": 1.9044,
-      "step": 463500
-    },
-    {
-      "epoch": 6.68,
-      "learning_rate": 5.2844287279506076e-05,
-      "loss": 1.9053,
-      "step": 464000
-    },
-    {
-      "epoch": 6.69,
-      "learning_rate": 5.279236053775334e-05,
-      "loss": 1.9006,
-      "step": 464500
-    },
-    {
-      "epoch": 6.69,
-      "learning_rate": 5.274043379600061e-05,
-      "loss": 1.9034,
-      "step": 465000
-    },
-    {
-      "epoch": 6.7,
-      "learning_rate": 5.2688507054247874e-05,
-      "loss": 1.9058,
-      "step": 465500
-    },
-    {
-      "epoch": 6.71,
-      "learning_rate": 5.263658031249513e-05,
-      "loss": 1.9032,
-      "step": 466000
-    },
-    {
-      "epoch": 6.71,
-      "learning_rate": 5.2584653570742404e-05,
-      "loss": 1.9006,
-      "step": 466500
-    },
-    {
-      "epoch": 6.72,
-      "learning_rate": 5.2532726828989665e-05,
-      "loss": 1.9026,
-      "step": 467000
-    },
-    {
-      "epoch": 6.73,
-      "learning_rate": 5.248090394072044e-05,
-      "loss": 1.904,
-      "step": 467500
-    },
-    {
-      "epoch": 6.74,
-      "learning_rate": 5.24289771989677e-05,
-      "loss": 1.8983,
-      "step": 468000
-    },
-    {
-      "epoch": 6.74,
-      "learning_rate": 5.237705045721496e-05,
-      "loss": 1.903,
-      "step": 468500
-    },
-    {
-      "epoch": 6.75,
-      "learning_rate": 5.232512371546223e-05,
-      "loss": 1.9042,
-      "step": 469000
-    },
-    {
-      "epoch": 6.76,
-      "learning_rate": 5.2273300827192996e-05,
-      "loss": 1.9001,
-      "step": 469500
-    },
-    {
-      "epoch": 6.77,
-      "learning_rate": 5.222147793892377e-05,
-      "loss": 1.9026,
-      "step": 470000
-    },
-    {
-      "epoch": 6.77,
-      "learning_rate": 5.216955119717103e-05,
-      "loss": 1.9003,
-      "step": 470500
-    },
-    {
-      "epoch": 6.78,
-      "learning_rate": 5.21176244554183e-05,
-      "loss": 1.9038,
-      "step": 471000
-    },
-    {
-      "epoch": 6.79,
-      "learning_rate": 5.206569771366556e-05,
-      "loss": 1.9011,
-      "step": 471500
-    },
-    {
-      "epoch": 6.79,
-      "learning_rate": 5.201377097191282e-05,
-      "loss": 1.9027,
-      "step": 472000
-    },
-    {
-      "epoch": 6.8,
-      "learning_rate": 5.19618442301601e-05,
-      "loss": 1.8988,
-      "step": 472500
-    },
-    {
-      "epoch": 6.81,
-      "learning_rate": 5.190991748840736e-05,
-      "loss": 1.9007,
-      "step": 473000
-    },
-    {
-      "epoch": 6.82,
-      "learning_rate": 5.185799074665462e-05,
-      "loss": 1.9063,
-      "step": 473500
-    },
-    {
-      "epoch": 6.82,
-      "learning_rate": 5.1806271711868904e-05,
-      "loss": 1.9045,
-      "step": 474000
-    },
-    {
-      "epoch": 6.83,
-      "learning_rate": 5.1754344970116166e-05,
-      "loss": 1.9047,
-      "step": 474500
-    },
-    {
-      "epoch": 6.84,
-      "learning_rate": 5.170241822836343e-05,
-      "loss": 1.8998,
-      "step": 475000
-    },
-    {
-      "epoch": 6.84,
-      "learning_rate": 5.1650491486610696e-05,
-      "loss": 1.9034,
-      "step": 475500
-    },
-    {
-      "epoch": 6.85,
-      "learning_rate": 5.159856474485796e-05,
-      "loss": 1.8985,
-      "step": 476000
-    },
-    {
-      "epoch": 6.86,
-      "learning_rate": 5.154663800310522e-05,
-      "loss": 1.9026,
-      "step": 476500
-    },
-    {
-      "epoch": 6.87,
-      "learning_rate": 5.1494711261352494e-05,
-      "loss": 1.8984,
-      "step": 477000
-    },
-    {
-      "epoch": 6.87,
-      "learning_rate": 5.144278451959975e-05,
-      "loss": 1.9015,
-      "step": 477500
-    },
-    {
-      "epoch": 6.88,
-      "learning_rate": 5.139096163133053e-05,
-      "loss": 1.9032,
-      "step": 478000
-    },
-    {
-      "epoch": 6.89,
-      "learning_rate": 5.133903488957778e-05,
-      "loss": 1.8992,
-      "step": 478500
-    },
-    {
-      "epoch": 6.89,
-      "learning_rate": 5.1287108147825045e-05,
-      "loss": 1.9035,
-      "step": 479000
-    },
-    {
-      "epoch": 6.9,
-      "learning_rate": 5.123518140607232e-05,
-      "loss": 1.9003,
-      "step": 479500
-    },
-    {
-      "epoch": 6.91,
-      "learning_rate": 5.118335851780308e-05,
-      "loss": 1.8944,
-      "step": 480000
-    },
-    {
-      "epoch": 6.92,
-      "learning_rate": 5.1131431776050354e-05,
-      "loss": 1.8998,
-      "step": 480500
-    },
-    {
-      "epoch": 6.92,
-      "learning_rate": 5.1079505034297616e-05,
-      "loss": 1.9034,
-      "step": 481000
-    },
-    {
-      "epoch": 6.93,
-      "learning_rate": 5.102757829254488e-05,
-      "loss": 1.9022,
-      "step": 481500
-    },
-    {
-      "epoch": 6.94,
-      "learning_rate": 5.097575540427565e-05,
-      "loss": 1.8972,
-      "step": 482000
-    },
-    {
-      "epoch": 6.95,
-      "learning_rate": 5.092382866252291e-05,
-      "loss": 1.9032,
-      "step": 482500
-    },
-    {
-      "epoch": 6.95,
-      "learning_rate": 5.087190192077018e-05,
-      "loss": 1.9014,
-      "step": 483000
-    },
-    {
-      "epoch": 6.96,
-      "learning_rate": 5.081997517901744e-05,
-      "loss": 1.8986,
-      "step": 483500
-    },
-    {
-      "epoch": 6.97,
-      "learning_rate": 5.0768152290748215e-05,
-      "loss": 1.8992,
-      "step": 484000
-    },
-    {
-      "epoch": 6.97,
-      "learning_rate": 5.0716225548995476e-05,
-      "loss": 1.9036,
-      "step": 484500
-    },
-    {
-      "epoch": 6.98,
-      "learning_rate": 5.066429880724275e-05,
-      "loss": 1.9018,
-      "step": 485000
-    },
-    {
-      "epoch": 6.99,
-      "learning_rate": 5.061237206549001e-05,
-      "loss": 1.8988,
-      "step": 485500
-    },
-    {
-      "epoch": 7.0,
-      "learning_rate": 5.0560549177220786e-05,
-      "loss": 1.8982,
-      "step": 486000
-    },
-    {
-      "epoch": 7.0,
-      "eval_accuracy": 0.6395200987544465,
-      "eval_loss": 1.757468581199646,
-      "eval_runtime": 647.437,
-      "eval_samples_per_second": 832.416,
-      "eval_steps_per_second": 34.684,
-      "step": 486311
-    },
-    {
-      "epoch": 7.0,
-      "learning_rate": 5.050862243546805e-05,
-      "loss": 1.8974,
-      "step": 486500
-    },
-    {
-      "epoch": 7.01,
-      "learning_rate": 5.045679954719882e-05,
-      "loss": 1.8964,
-      "step": 487000
-    },
-    {
-      "epoch": 7.02,
-      "learning_rate": 5.040487280544608e-05,
-      "loss": 1.8951,
-      "step": 487500
-    },
-    {
-      "epoch": 7.02,
-      "learning_rate": 5.035294606369334e-05,
-      "loss": 1.8956,
-      "step": 488000
-    },
-    {
-      "epoch": 7.03,
-      "learning_rate": 5.030101932194061e-05,
-      "loss": 1.8932,
-      "step": 488500
-    },
-    {
-      "epoch": 7.04,
-      "learning_rate": 5.024909258018787e-05,
-      "loss": 1.8955,
-      "step": 489000
-    },
-    {
-      "epoch": 7.05,
-      "learning_rate": 5.0197165838435135e-05,
-      "loss": 1.8931,
-      "step": 489500
-    },
-    {
-      "epoch": 7.05,
-      "learning_rate": 5.01452390966824e-05,
-      "loss": 1.8942,
-      "step": 490000
-    },
-    {
-      "epoch": 7.06,
-      "learning_rate": 5.0093312354929664e-05,
-      "loss": 1.8929,
-      "step": 490500
-    },
-    {
-      "epoch": 7.07,
-      "learning_rate": 5.004148946666044e-05,
-      "loss": 1.8949,
-      "step": 491000
-    },
-    {
-      "epoch": 7.07,
-      "learning_rate": 4.99895627249077e-05,
-      "loss": 1.9012,
-      "step": 491500
-    },
-    {
-      "epoch": 7.08,
-      "learning_rate": 4.993763598315497e-05,
-      "loss": 1.8931,
-      "step": 492000
-    },
-    {
-      "epoch": 7.09,
-      "learning_rate": 4.9885709241402236e-05,
-      "loss": 1.8933,
-      "step": 492500
-    },
-    {
-      "epoch": 7.1,
-      "learning_rate": 4.9833886353133e-05,
-      "loss": 1.8939,
-      "step": 493000
-    },
-    {
-      "epoch": 7.1,
-      "learning_rate": 4.978195961138027e-05,
-      "loss": 1.8947,
-      "step": 493500
-    },
-    {
-      "epoch": 7.11,
-      "learning_rate": 4.973003286962753e-05,
-      "loss": 1.8957,
-      "step": 494000
-    },
-    {
-      "epoch": 7.12,
-      "learning_rate": 4.967810612787479e-05,
-      "loss": 1.8973,
-      "step": 494500
-    },
-    {
-      "epoch": 7.13,
-      "learning_rate": 4.9626283239605566e-05,
-      "loss": 1.8952,
-      "step": 495000
-    },
-    {
-      "epoch": 7.13,
-      "learning_rate": 4.957435649785283e-05,
-      "loss": 1.8993,
-      "step": 495500
-    },
-    {
-      "epoch": 7.14,
-      "learning_rate": 4.9522429756100096e-05,
-      "loss": 1.8981,
-      "step": 496000
-    },
-    {
-      "epoch": 7.15,
-      "learning_rate": 4.9470503014347364e-05,
-      "loss": 1.8928,
-      "step": 496500
-    },
-    {
-      "epoch": 7.15,
-      "learning_rate": 4.941868012607813e-05,
-      "loss": 1.8949,
-      "step": 497000
-    },
-    {
-      "epoch": 7.16,
-      "learning_rate": 4.93668572378089e-05,
-      "loss": 1.8945,
-      "step": 497500
-    },
-    {
-      "epoch": 7.17,
-      "learning_rate": 4.9314930496056165e-05,
-      "loss": 1.8971,
-      "step": 498000
-    },
-    {
-      "epoch": 7.18,
-      "learning_rate": 4.926300375430343e-05,
-      "loss": 1.8924,
-      "step": 498500
-    },
-    {
-      "epoch": 7.18,
-      "learning_rate": 4.9211077012550695e-05,
-      "loss": 1.894,
-      "step": 499000
-    },
-    {
-      "epoch": 7.19,
-      "learning_rate": 4.915915027079796e-05,
-      "loss": 1.8934,
-      "step": 499500
-    },
-    {
-      "epoch": 7.2,
-      "learning_rate": 4.9107223529045225e-05,
-      "loss": 1.8972,
-      "step": 500000
-    },
-    {
-      "epoch": 7.2,
-      "learning_rate": 4.9055296787292486e-05,
-      "loss": 1.8903,
-      "step": 500500
-    },
-    {
-      "epoch": 7.21,
-      "learning_rate": 4.9003370045539754e-05,
-      "loss": 1.8908,
-      "step": 501000
-    },
-    {
-      "epoch": 7.22,
-      "learning_rate": 4.895154715727053e-05,
-      "loss": 1.8885,
-      "step": 501500
-    },
-    {
-      "epoch": 7.23,
-      "learning_rate": 4.889962041551779e-05,
-      "loss": 1.89,
-      "step": 502000
-    },
-    {
-      "epoch": 7.23,
-      "learning_rate": 4.884769367376506e-05,
-      "loss": 1.8922,
-      "step": 502500
-    },
-    {
-      "epoch": 7.24,
-      "learning_rate": 4.879576693201232e-05,
-      "loss": 1.896,
-      "step": 503000
-    },
-    {
-      "epoch": 7.25,
-      "learning_rate": 4.874394404374309e-05,
-      "loss": 1.8955,
-      "step": 503500
-    },
-    {
-      "epoch": 7.25,
-      "learning_rate": 4.869201730199035e-05,
-      "loss": 1.893,
-      "step": 504000
-    },
-    {
-      "epoch": 7.26,
-      "learning_rate": 4.8640090560237615e-05,
-      "loss": 1.8923,
-      "step": 504500
-    },
-    {
-      "epoch": 7.27,
-      "learning_rate": 4.858816381848488e-05,
-      "loss": 1.8926,
-      "step": 505000
-    },
-    {
-      "epoch": 7.28,
-      "learning_rate": 4.853634093021565e-05,
-      "loss": 1.8888,
-      "step": 505500
-    },
-    {
-      "epoch": 7.28,
-      "learning_rate": 4.848441418846292e-05,
-      "loss": 1.8936,
-      "step": 506000
-    },
-    {
-      "epoch": 7.29,
-      "learning_rate": 4.8432487446710186e-05,
-      "loss": 1.8947,
-      "step": 506500
-    },
-    {
-      "epoch": 7.3,
-      "learning_rate": 4.838056070495745e-05,
-      "loss": 1.8902,
-      "step": 507000
-    },
-    {
-      "epoch": 7.3,
-      "learning_rate": 4.8328841670171725e-05,
-      "loss": 1.8917,
-      "step": 507500
-    },
-    {
-      "epoch": 7.31,
-      "learning_rate": 4.827691492841899e-05,
-      "loss": 1.8927,
-      "step": 508000
-    },
-    {
-      "epoch": 7.32,
-      "learning_rate": 4.8224988186666255e-05,
-      "loss": 1.8886,
-      "step": 508500
-    },
-    {
-      "epoch": 7.33,
-      "learning_rate": 4.8173061444913516e-05,
-      "loss": 1.8931,
-      "step": 509000
-    },
-    {
-      "epoch": 7.33,
-      "learning_rate": 4.812113470316078e-05,
-      "loss": 1.8912,
-      "step": 509500
-    },
-    {
-      "epoch": 7.34,
-      "learning_rate": 4.806931181489156e-05,
-      "loss": 1.8961,
-      "step": 510000
-    },
-    {
-      "epoch": 7.35,
-      "learning_rate": 4.801738507313881e-05,
-      "loss": 1.8916,
-      "step": 510500
-    },
-    {
-      "epoch": 7.36,
-      "learning_rate": 4.796545833138608e-05,
-      "loss": 1.8948,
-      "step": 511000
-    },
-    {
-      "epoch": 7.36,
-      "learning_rate": 4.791353158963335e-05,
-      "loss": 1.8891,
-      "step": 511500
-    },
-    {
-      "epoch": 7.37,
-      "learning_rate": 4.7861708701364115e-05,
-      "loss": 1.8916,
-      "step": 512000
-    },
-    {
-      "epoch": 7.38,
-      "learning_rate": 4.7809781959611384e-05,
-      "loss": 1.8918,
-      "step": 512500
-    },
-    {
-      "epoch": 7.38,
-      "learning_rate": 4.775785521785865e-05,
-      "loss": 1.892,
-      "step": 513000
-    },
-    {
-      "epoch": 7.39,
-      "learning_rate": 4.770592847610591e-05,
-      "loss": 1.8879,
-      "step": 513500
-    },
-    {
-      "epoch": 7.4,
-      "learning_rate": 4.7654001734353175e-05,
-      "loss": 1.8884,
-      "step": 514000
-    },
-    {
-      "epoch": 7.41,
-      "learning_rate": 4.760207499260044e-05,
-      "loss": 1.8874,
-      "step": 514500
-    },
-    {
-      "epoch": 7.41,
-      "learning_rate": 4.7550148250847705e-05,
-      "loss": 1.8866,
-      "step": 515000
-    },
-    {
-      "epoch": 7.42,
-      "learning_rate": 4.749822150909497e-05,
-      "loss": 1.8935,
-      "step": 515500
-    },
-    {
-      "epoch": 7.43,
-      "learning_rate": 4.7446398620825746e-05,
-      "loss": 1.8901,
-      "step": 516000
-    },
-    {
-      "epoch": 7.43,
-      "learning_rate": 4.739447187907301e-05,
-      "loss": 1.8954,
-      "step": 516500
-    },
-    {
-      "epoch": 7.44,
-      "learning_rate": 4.734254513732027e-05,
-      "loss": 1.8914,
-      "step": 517000
-    },
-    {
-      "epoch": 7.45,
-      "learning_rate": 4.729061839556754e-05,
-      "loss": 1.89,
-      "step": 517500
-    },
-    {
-      "epoch": 7.46,
-      "learning_rate": 4.7238795507298304e-05,
-      "loss": 1.8898,
-      "step": 518000
-    },
-    {
-      "epoch": 7.46,
-      "learning_rate": 4.718686876554557e-05,
-      "loss": 1.8849,
-      "step": 518500
-    },
-    {
-      "epoch": 7.47,
-      "learning_rate": 4.7134942023792833e-05,
-      "loss": 1.8887,
-      "step": 519000
-    },
-    {
-      "epoch": 7.48,
-      "learning_rate": 4.70830152820401e-05,
-      "loss": 1.8914,
-      "step": 519500
-    },
-    {
-      "epoch": 7.48,
-      "learning_rate": 4.703119239377087e-05,
-      "loss": 1.8886,
-      "step": 520000
-    },
-    {
-      "epoch": 7.49,
-      "learning_rate": 4.6979265652018136e-05,
-      "loss": 1.8832,
-      "step": 520500
-    },
-    {
-      "epoch": 7.5,
-      "learning_rate": 4.69273389102654e-05,
-      "loss": 1.8941,
-      "step": 521000
-    },
-    {
-      "epoch": 7.51,
-      "learning_rate": 4.6875412168512666e-05,
-      "loss": 1.8879,
-      "step": 521500
-    },
-    {
-      "epoch": 7.51,
-      "learning_rate": 4.682358928024343e-05,
-      "loss": 1.8884,
-      "step": 522000
-    },
-    {
-      "epoch": 7.52,
-      "learning_rate": 4.67716625384907e-05,
-      "loss": 1.8884,
-      "step": 522500
-    },
-    {
-      "epoch": 7.53,
-      "learning_rate": 4.671973579673796e-05,
-      "loss": 1.8877,
-      "step": 523000
-    },
-    {
-      "epoch": 7.54,
-      "learning_rate": 4.666780905498523e-05,
-      "loss": 1.8903,
-      "step": 523500
-    },
-    {
-      "epoch": 7.54,
-      "learning_rate": 4.6615986166716e-05,
-      "loss": 1.8854,
-      "step": 524000
-    },
-    {
-      "epoch": 7.55,
-      "learning_rate": 4.6564059424963265e-05,
-      "loss": 1.8902,
-      "step": 524500
-    },
-    {
-      "epoch": 7.56,
-      "learning_rate": 4.6512132683210526e-05,
-      "loss": 1.8871,
-      "step": 525000
-    },
-    {
-      "epoch": 7.56,
-      "learning_rate": 4.64603097949413e-05,
-      "loss": 1.8913,
-      "step": 525500
-    },
-    {
-      "epoch": 7.57,
-      "learning_rate": 4.640838305318856e-05,
-      "loss": 1.8844,
-      "step": 526000
-    },
-    {
-      "epoch": 7.58,
-      "learning_rate": 4.635645631143583e-05,
-      "loss": 1.8863,
-      "step": 526500
-    },
-    {
-      "epoch": 7.59,
-      "learning_rate": 4.630452956968309e-05,
-      "loss": 1.8858,
-      "step": 527000
-    },
-    {
-      "epoch": 7.59,
-      "learning_rate": 4.6252706681413864e-05,
-      "loss": 1.8871,
-      "step": 527500
-    },
-    {
-      "epoch": 7.6,
-      "learning_rate": 4.6200779939661125e-05,
-      "loss": 1.8851,
-      "step": 528000
-    },
-    {
-      "epoch": 7.61,
-      "learning_rate": 4.6148853197908394e-05,
-      "loss": 1.8875,
-      "step": 528500
-    },
-    {
-      "epoch": 7.61,
-      "learning_rate": 4.609692645615566e-05,
-      "loss": 1.8862,
-      "step": 529000
-    },
-    {
-      "epoch": 7.62,
-      "learning_rate": 4.604499971440292e-05,
-      "loss": 1.8868,
-      "step": 529500
-    },
-    {
-      "epoch": 7.63,
-      "learning_rate": 4.5993072972650185e-05,
-      "loss": 1.8885,
-      "step": 530000
-    },
-    {
-      "epoch": 7.64,
-      "learning_rate": 4.594114623089745e-05,
-      "loss": 1.8867,
-      "step": 530500
-    },
-    {
-      "epoch": 7.64,
-      "learning_rate": 4.588932334262822e-05,
-      "loss": 1.8873,
-      "step": 531000
-    },
-    {
-      "epoch": 7.65,
-      "learning_rate": 4.583739660087549e-05,
-      "loss": 1.8849,
-      "step": 531500
-    },
-    {
-      "epoch": 7.66,
-      "learning_rate": 4.5785469859122756e-05,
-      "loss": 1.8829,
-      "step": 532000
-    },
-    {
-      "epoch": 7.66,
-      "learning_rate": 4.573354311737002e-05,
-      "loss": 1.8851,
-      "step": 532500
-    },
-    {
-      "epoch": 7.67,
-      "learning_rate": 4.568161637561728e-05,
-      "loss": 1.8881,
-      "step": 533000
-    },
-    {
-      "epoch": 7.68,
-      "learning_rate": 4.562979348734805e-05,
-      "loss": 1.8836,
-      "step": 533500
-    },
-    {
-      "epoch": 7.69,
-      "learning_rate": 4.5577866745595314e-05,
-      "loss": 1.8855,
-      "step": 534000
-    },
-    {
-      "epoch": 7.69,
-      "learning_rate": 4.552594000384258e-05,
-      "loss": 1.8858,
-      "step": 534500
-    },
-    {
-      "epoch": 7.7,
-      "learning_rate": 4.547401326208985e-05,
-      "loss": 1.89,
-      "step": 535000
-    },
-    {
-      "epoch": 7.71,
-      "learning_rate": 4.542208652033711e-05,
-      "loss": 1.8891,
-      "step": 535500
-    },
-    {
-      "epoch": 7.72,
-      "learning_rate": 4.537015977858437e-05,
-      "loss": 1.8859,
-      "step": 536000
-    },
-    {
-      "epoch": 7.72,
-      "learning_rate": 4.531823303683164e-05,
-      "loss": 1.8885,
-      "step": 536500
-    },
-    {
-      "epoch": 7.73,
-      "learning_rate": 4.52663062950789e-05,
-      "loss": 1.886,
-      "step": 537000
-    },
-    {
-      "epoch": 7.74,
-      "learning_rate": 4.5214483406809676e-05,
-      "loss": 1.8879,
-      "step": 537500
-    },
-    {
-      "epoch": 7.74,
-      "learning_rate": 4.516266051854044e-05,
-      "loss": 1.8859,
-      "step": 538000
-    },
-    {
-      "epoch": 7.75,
-      "learning_rate": 4.511073377678771e-05,
-      "loss": 1.8838,
-      "step": 538500
-    },
-    {
-      "epoch": 7.76,
-      "learning_rate": 4.505880703503497e-05,
-      "loss": 1.883,
-      "step": 539000
-    },
-    {
-      "epoch": 7.77,
-      "learning_rate": 4.500688029328224e-05,
-      "loss": 1.8858,
-      "step": 539500
-    },
-    {
-      "epoch": 7.77,
-      "learning_rate": 4.49549535515295e-05,
-      "loss": 1.886,
-      "step": 540000
-    },
-    {
-      "epoch": 7.78,
-      "learning_rate": 4.490302680977677e-05,
-      "loss": 1.8845,
-      "step": 540500
-    },
-    {
-      "epoch": 7.79,
-      "learning_rate": 4.485110006802403e-05,
-      "loss": 1.8904,
-      "step": 541000
-    },
-    {
-      "epoch": 7.79,
-      "learning_rate": 4.47991733262713e-05,
-      "loss": 1.8853,
-      "step": 541500
-    },
-    {
-      "epoch": 7.8,
-      "learning_rate": 4.4747350438002066e-05,
-      "loss": 1.8849,
-      "step": 542000
-    },
-    {
-      "epoch": 7.81,
-      "learning_rate": 4.4695423696249335e-05,
-      "loss": 1.8841,
-      "step": 542500
-    },
-    {
-      "epoch": 7.82,
-      "learning_rate": 4.46434969544966e-05,
-      "loss": 1.886,
-      "step": 543000
-    },
-    {
-      "epoch": 7.82,
-      "learning_rate": 4.459157021274386e-05,
-      "loss": 1.8844,
-      "step": 543500
-    },
-    {
-      "epoch": 7.83,
-      "learning_rate": 4.453974732447464e-05,
-      "loss": 1.8831,
-      "step": 544000
-    },
-    {
-      "epoch": 7.84,
-      "learning_rate": 4.44878205827219e-05,
-      "loss": 1.8854,
-      "step": 544500
-    },
-    {
-      "epoch": 7.84,
-      "learning_rate": 4.443589384096916e-05,
-      "loss": 1.8829,
-      "step": 545000
-    },
-    {
-      "epoch": 7.85,
-      "learning_rate": 4.438396709921643e-05,
-      "loss": 1.8865,
-      "step": 545500
-    },
-    {
-      "epoch": 7.86,
-      "learning_rate": 4.4332144210947195e-05,
-      "loss": 1.8806,
-      "step": 546000
-    },
-    {
-      "epoch": 7.87,
-      "learning_rate": 4.428021746919446e-05,
-      "loss": 1.8838,
-      "step": 546500
-    },
-    {
-      "epoch": 7.87,
-      "learning_rate": 4.422829072744173e-05,
-      "loss": 1.8853,
-      "step": 547000
-    },
-    {
-      "epoch": 7.88,
-      "learning_rate": 4.417636398568899e-05,
-      "loss": 1.8821,
-      "step": 547500
-    },
-    {
-      "epoch": 7.89,
-      "learning_rate": 4.4124541097419766e-05,
-      "loss": 1.8817,
-      "step": 548000
-    },
-    {
-      "epoch": 7.9,
-      "learning_rate": 4.407271820915053e-05,
-      "loss": 1.8797,
-      "step": 548500
-    },
-    {
-      "epoch": 7.9,
-      "learning_rate": 4.40207914673978e-05,
-      "loss": 1.8848,
-      "step": 549000
-    },
-    {
-      "epoch": 7.91,
-      "learning_rate": 4.396886472564506e-05,
-      "loss": 1.8852,
-      "step": 549500
-    },
-    {
-      "epoch": 7.92,
-      "learning_rate": 4.3916937983892324e-05,
-      "loss": 1.8805,
-      "step": 550000
-    },
-    {
-      "epoch": 7.92,
-      "learning_rate": 4.386501124213959e-05,
-      "loss": 1.8832,
-      "step": 550500
-    },
-    {
-      "epoch": 7.93,
-      "learning_rate": 4.381308450038686e-05,
-      "loss": 1.8835,
-      "step": 551000
-    },
-    {
-      "epoch": 7.94,
-      "learning_rate": 4.376115775863412e-05,
-      "loss": 1.8816,
-      "step": 551500
-    },
-    {
-      "epoch": 7.95,
-      "learning_rate": 4.370923101688138e-05,
-      "loss": 1.882,
-      "step": 552000
-    },
-    {
-      "epoch": 7.95,
-      "learning_rate": 4.3657408128612156e-05,
-      "loss": 1.8875,
-      "step": 552500
-    },
-    {
-      "epoch": 7.96,
-      "learning_rate": 4.360548138685942e-05,
-      "loss": 1.881,
-      "step": 553000
-    },
-    {
-      "epoch": 7.97,
-      "learning_rate": 4.3553554645106686e-05,
-      "loss": 1.8777,
-      "step": 553500
-    },
-    {
-      "epoch": 7.97,
-      "learning_rate": 4.3501627903353954e-05,
-      "loss": 1.8788,
-      "step": 554000
-    },
-    {
-      "epoch": 7.98,
-      "learning_rate": 4.344980501508472e-05,
-      "loss": 1.886,
-      "step": 554500
-    },
-    {
-      "epoch": 7.99,
-      "learning_rate": 4.339787827333199e-05,
-      "loss": 1.8836,
-      "step": 555000
-    },
-    {
-      "epoch": 8.0,
-      "learning_rate": 4.334595153157925e-05,
-      "loss": 1.8808,
-      "step": 555500
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.6421464828448523,
-      "eval_loss": 1.741304636001587,
-      "eval_runtime": 646.83,
-      "eval_samples_per_second": 833.197,
-      "eval_steps_per_second": 34.717,
-      "step": 555784
-    },
-    {
-      "epoch": 8.0,
-      "learning_rate": 4.3294128643310023e-05,
-      "loss": 1.8816,
-      "step": 556000
-    },
-    {
-      "epoch": 8.01,
-      "learning_rate": 4.3242201901557285e-05,
-      "loss": 1.8744,
-      "step": 556500
-    },
-    {
-      "epoch": 8.02,
-      "learning_rate": 4.3190275159804546e-05,
-      "loss": 1.8789,
-      "step": 557000
-    },
-    {
-      "epoch": 8.02,
-      "learning_rate": 4.3138348418051815e-05,
-      "loss": 1.8776,
-      "step": 557500
-    },
-    {
-      "epoch": 8.03,
-      "learning_rate": 4.3086421676299076e-05,
-      "loss": 1.8818,
-      "step": 558000
-    },
-    {
-      "epoch": 8.04,
-      "learning_rate": 4.3034494934546345e-05,
-      "loss": 1.8787,
-      "step": 558500
-    },
-    {
-      "epoch": 8.05,
-      "learning_rate": 4.298256819279361e-05,
-      "loss": 1.8769,
-      "step": 559000
-    },
-    {
-      "epoch": 8.05,
-      "learning_rate": 4.2930641451040874e-05,
-      "loss": 1.8796,
-      "step": 559500
-    },
-    {
-      "epoch": 8.06,
-      "learning_rate": 4.287881856277165e-05,
-      "loss": 1.8799,
-      "step": 560000
-    },
-    {
-      "epoch": 8.07,
-      "learning_rate": 4.282689182101891e-05,
-      "loss": 1.8759,
-      "step": 560500
-    },
-    {
-      "epoch": 8.08,
-      "learning_rate": 4.277496507926617e-05,
-      "loss": 1.8762,
-      "step": 561000
-    },
-    {
-      "epoch": 8.08,
-      "learning_rate": 4.272303833751344e-05,
-      "loss": 1.8775,
-      "step": 561500
-    },
-    {
-      "epoch": 8.09,
-      "learning_rate": 4.2671215449244205e-05,
-      "loss": 1.8778,
-      "step": 562000
-    },
-    {
-      "epoch": 8.1,
-      "learning_rate": 4.261928870749147e-05,
-      "loss": 1.8753,
-      "step": 562500
-    },
-    {
-      "epoch": 8.1,
-      "learning_rate": 4.256746581922224e-05,
-      "loss": 1.8799,
-      "step": 563000
-    },
-    {
-      "epoch": 8.11,
-      "learning_rate": 4.251553907746951e-05,
-      "loss": 1.8741,
-      "step": 563500
-    },
-    {
-      "epoch": 8.12,
-      "learning_rate": 4.2463612335716776e-05,
-      "loss": 1.8765,
-      "step": 564000
-    },
-    {
-      "epoch": 8.13,
-      "learning_rate": 4.241168559396404e-05,
-      "loss": 1.8734,
-      "step": 564500
-    },
-    {
-      "epoch": 8.13,
-      "learning_rate": 4.23597588522113e-05,
-      "loss": 1.8799,
-      "step": 565000
-    },
-    {
-      "epoch": 8.14,
-      "learning_rate": 4.230783211045857e-05,
-      "loss": 1.8719,
-      "step": 565500
-    },
-    {
-      "epoch": 8.15,
-      "learning_rate": 4.2255905368705836e-05,
-      "loss": 1.874,
-      "step": 566000
-    },
-    {
-      "epoch": 8.15,
-      "learning_rate": 4.22039786269531e-05,
-      "loss": 1.879,
-      "step": 566500
-    },
-    {
-      "epoch": 8.16,
-      "learning_rate": 4.215215573868387e-05,
-      "loss": 1.8821,
-      "step": 567000
-    },
-    {
-      "epoch": 8.17,
-      "learning_rate": 4.210022899693113e-05,
-      "loss": 1.8782,
-      "step": 567500
-    },
-    {
-      "epoch": 8.18,
-      "learning_rate": 4.204830225517839e-05,
-      "loss": 1.8776,
-      "step": 568000
-    },
-    {
-      "epoch": 8.18,
-      "learning_rate": 4.199637551342566e-05,
-      "loss": 1.8789,
-      "step": 568500
-    },
-    {
-      "epoch": 8.19,
-      "learning_rate": 4.194455262515643e-05,
-      "loss": 1.874,
-      "step": 569000
-    },
-    {
-      "epoch": 8.2,
-      "learning_rate": 4.18927297368872e-05,
-      "loss": 1.8778,
-      "step": 569500
-    },
-    {
-      "epoch": 8.2,
-      "learning_rate": 4.184080299513446e-05,
-      "loss": 1.8758,
-      "step": 570000
-    },
-    {
-      "epoch": 8.21,
-      "learning_rate": 4.178887625338173e-05,
-      "loss": 1.8775,
-      "step": 570500
-    },
-    {
-      "epoch": 8.22,
-      "learning_rate": 4.1736949511629e-05,
-      "loss": 1.8754,
-      "step": 571000
-    },
-    {
-      "epoch": 8.23,
-      "learning_rate": 4.168502276987626e-05,
-      "loss": 1.8767,
-      "step": 571500
-    },
-    {
-      "epoch": 8.23,
-      "learning_rate": 4.163309602812352e-05,
-      "loss": 1.873,
-      "step": 572000
-    },
-    {
-      "epoch": 8.24,
-      "learning_rate": 4.158116928637079e-05,
-      "loss": 1.8753,
-      "step": 572500
-    },
-    {
-      "epoch": 8.25,
-      "learning_rate": 4.152924254461805e-05,
-      "loss": 1.8738,
-      "step": 573000
-    },
-    {
-      "epoch": 8.26,
-      "learning_rate": 4.1477419656348825e-05,
-      "loss": 1.8776,
-      "step": 573500
-    },
-    {
-      "epoch": 8.26,
-      "learning_rate": 4.142559676807959e-05,
-      "loss": 1.8794,
-      "step": 574000
-    },
-    {
-      "epoch": 8.27,
-      "learning_rate": 4.137367002632686e-05,
-      "loss": 1.8763,
-      "step": 574500
-    },
-    {
-      "epoch": 8.28,
-      "learning_rate": 4.132174328457413e-05,
-      "loss": 1.8744,
-      "step": 575000
-    },
-    {
-      "epoch": 8.28,
-      "learning_rate": 4.126981654282139e-05,
-      "loss": 1.8733,
-      "step": 575500
-    },
-    {
-      "epoch": 8.29,
-      "learning_rate": 4.121788980106866e-05,
-      "loss": 1.8696,
-      "step": 576000
-    },
-    {
-      "epoch": 8.3,
-      "learning_rate": 4.116596305931592e-05,
-      "loss": 1.876,
-      "step": 576500
-    },
-    {
-      "epoch": 8.31,
-      "learning_rate": 4.111403631756318e-05,
-      "loss": 1.8778,
-      "step": 577000
-    },
-    {
-      "epoch": 8.31,
-      "learning_rate": 4.106210957581045e-05,
-      "loss": 1.8761,
-      "step": 577500
-    },
-    {
-      "epoch": 8.32,
-      "learning_rate": 4.1010286687541215e-05,
-      "loss": 1.8761,
-      "step": 578000
-    },
-    {
-      "epoch": 8.33,
-      "learning_rate": 4.095835994578848e-05,
-      "loss": 1.8759,
-      "step": 578500
-    },
-    {
-      "epoch": 8.33,
-      "learning_rate": 4.090643320403575e-05,
-      "loss": 1.8725,
-      "step": 579000
-    },
-    {
-      "epoch": 8.34,
-      "learning_rate": 4.085450646228301e-05,
-      "loss": 1.8773,
-      "step": 579500
-    },
-    {
-      "epoch": 8.35,
-      "learning_rate": 4.0802683574013786e-05,
-      "loss": 1.8752,
-      "step": 580000
-    },
-    {
-      "epoch": 8.36,
-      "learning_rate": 4.075075683226105e-05,
-      "loss": 1.8776,
-      "step": 580500
-    },
-    {
-      "epoch": 8.36,
-      "learning_rate": 4.069883009050831e-05,
-      "loss": 1.8756,
-      "step": 581000
-    },
-    {
-      "epoch": 8.37,
-      "learning_rate": 4.064690334875558e-05,
-      "loss": 1.8744,
-      "step": 581500
-    },
-    {
-      "epoch": 8.38,
-      "learning_rate": 4.0594976607002846e-05,
-      "loss": 1.8796,
-      "step": 582000
-    },
-    {
-      "epoch": 8.38,
-      "learning_rate": 4.054304986525011e-05,
-      "loss": 1.8739,
-      "step": 582500
-    },
-    {
-      "epoch": 8.39,
-      "learning_rate": 4.049112312349737e-05,
-      "loss": 1.8727,
-      "step": 583000
-    },
-    {
-      "epoch": 8.4,
-      "learning_rate": 4.043919638174464e-05,
-      "loss": 1.8734,
-      "step": 583500
-    },
-    {
-      "epoch": 8.41,
-      "learning_rate": 4.03873734934754e-05,
-      "loss": 1.8798,
-      "step": 584000
-    },
-    {
-      "epoch": 8.41,
-      "learning_rate": 4.033544675172267e-05,
-      "loss": 1.8713,
-      "step": 584500
-    },
-    {
-      "epoch": 8.42,
-      "learning_rate": 4.028352000996994e-05,
-      "loss": 1.8746,
-      "step": 585000
-    },
-    {
-      "epoch": 8.43,
-      "learning_rate": 4.02315932682172e-05,
-      "loss": 1.8737,
-      "step": 585500
-    },
-    {
-      "epoch": 8.43,
-      "learning_rate": 4.0179770379947974e-05,
-      "loss": 1.8744,
-      "step": 586000
-    },
-    {
-      "epoch": 8.44,
-      "learning_rate": 4.0127843638195236e-05,
-      "loss": 1.875,
-      "step": 586500
-    },
-    {
-      "epoch": 8.45,
-      "learning_rate": 4.00759168964425e-05,
-      "loss": 1.8761,
-      "step": 587000
-    },
-    {
-      "epoch": 8.46,
-      "learning_rate": 4.0023990154689766e-05,
-      "loss": 1.8725,
-      "step": 587500
-    },
-    {
-      "epoch": 8.46,
-      "learning_rate": 3.997216726642053e-05,
-      "loss": 1.8763,
-      "step": 588000
-    },
-    {
-      "epoch": 8.47,
-      "learning_rate": 3.99202405246678e-05,
-      "loss": 1.8745,
-      "step": 588500
-    },
-    {
-      "epoch": 8.48,
-      "learning_rate": 3.986831378291507e-05,
-      "loss": 1.8748,
-      "step": 589000
-    },
-    {
-      "epoch": 8.49,
-      "learning_rate": 3.981638704116233e-05,
-      "loss": 1.8732,
-      "step": 589500
-    },
-    {
-      "epoch": 8.49,
-      "learning_rate": 3.97645641528931e-05,
-      "loss": 1.875,
-      "step": 590000
-    },
-    {
-      "epoch": 8.5,
-      "learning_rate": 3.971274126462387e-05,
-      "loss": 1.8705,
-      "step": 590500
-    },
-    {
-      "epoch": 8.51,
-      "learning_rate": 3.966081452287114e-05,
-      "loss": 1.8717,
-      "step": 591000
-    },
-    {
-      "epoch": 8.51,
-      "learning_rate": 3.96088877811184e-05,
-      "loss": 1.8737,
-      "step": 591500
-    },
-    {
-      "epoch": 8.52,
-      "learning_rate": 3.955696103936567e-05,
-      "loss": 1.8713,
-      "step": 592000
-    },
-    {
-      "epoch": 8.53,
-      "learning_rate": 3.950503429761293e-05,
-      "loss": 1.8764,
-      "step": 592500
-    },
-    {
-      "epoch": 8.54,
-      "learning_rate": 3.94531075558602e-05,
-      "loss": 1.8714,
-      "step": 593000
-    },
-    {
-      "epoch": 8.54,
-      "learning_rate": 3.940118081410746e-05,
-      "loss": 1.8711,
-      "step": 593500
-    },
-    {
-      "epoch": 8.55,
-      "learning_rate": 3.934925407235473e-05,
-      "loss": 1.871,
-      "step": 594000
-    },
-    {
-      "epoch": 8.56,
-      "learning_rate": 3.929743118408549e-05,
-      "loss": 1.8695,
-      "step": 594500
-    },
-    {
-      "epoch": 8.56,
-      "learning_rate": 3.924550444233276e-05,
-      "loss": 1.8734,
-      "step": 595000
-    },
-    {
-      "epoch": 8.57,
-      "learning_rate": 3.919357770058002e-05,
-      "loss": 1.8704,
-      "step": 595500
-    },
-    {
-      "epoch": 8.58,
-      "learning_rate": 3.9141650958827285e-05,
-      "loss": 1.8703,
-      "step": 596000
-    },
-    {
-      "epoch": 8.59,
-      "learning_rate": 3.908972421707455e-05,
-      "loss": 1.8733,
-      "step": 596500
-    },
-    {
-      "epoch": 8.59,
-      "learning_rate": 3.903779747532182e-05,
-      "loss": 1.8698,
-      "step": 597000
-    },
-    {
-      "epoch": 8.6,
-      "learning_rate": 3.898587073356908e-05,
-      "loss": 1.8741,
-      "step": 597500
-    },
-    {
-      "epoch": 8.61,
-      "learning_rate": 3.8933943991816344e-05,
-      "loss": 1.8723,
-      "step": 598000
-    },
-    {
-      "epoch": 8.61,
-      "learning_rate": 3.888212110354712e-05,
-      "loss": 1.8735,
-      "step": 598500
-    },
-    {
-      "epoch": 8.62,
-      "learning_rate": 3.883019436179438e-05,
-      "loss": 1.8716,
-      "step": 599000
-    },
-    {
-      "epoch": 8.63,
-      "learning_rate": 3.877826762004165e-05,
-      "loss": 1.8689,
-      "step": 599500
-    },
-    {
-      "epoch": 8.64,
-      "learning_rate": 3.8726340878288915e-05,
-      "loss": 1.869,
-      "step": 600000
-    },
-    {
-      "epoch": 8.64,
-      "learning_rate": 3.867451799001968e-05,
-      "loss": 1.8711,
-      "step": 600500
-    },
-    {
-      "epoch": 8.65,
-      "learning_rate": 3.862259124826695e-05,
-      "loss": 1.8725,
-      "step": 601000
-    },
-    {
-      "epoch": 8.66,
-      "learning_rate": 3.857066450651421e-05,
-      "loss": 1.87,
-      "step": 601500
-    },
-    {
-      "epoch": 8.67,
-      "learning_rate": 3.851873776476147e-05,
-      "loss": 1.8694,
-      "step": 602000
-    },
-    {
-      "epoch": 8.67,
-      "learning_rate": 3.846681102300874e-05,
-      "loss": 1.8677,
-      "step": 602500
-    },
-    {
-      "epoch": 8.68,
-      "learning_rate": 3.841498813473951e-05,
-      "loss": 1.8681,
-      "step": 603000
-    },
-    {
-      "epoch": 8.69,
-      "learning_rate": 3.8363061392986776e-05,
-      "loss": 1.8718,
-      "step": 603500
-    },
-    {
-      "epoch": 8.69,
-      "learning_rate": 3.8311134651234044e-05,
-      "loss": 1.8694,
-      "step": 604000
-    },
-    {
-      "epoch": 8.7,
-      "learning_rate": 3.8259207909481305e-05,
-      "loss": 1.8731,
-      "step": 604500
-    },
-    {
-      "epoch": 8.71,
-      "learning_rate": 3.820738502121208e-05,
-      "loss": 1.8699,
-      "step": 605000
-    },
-    {
-      "epoch": 8.72,
-      "learning_rate": 3.815545827945934e-05,
-      "loss": 1.8737,
-      "step": 605500
-    },
-    {
-      "epoch": 8.72,
-      "learning_rate": 3.81035315377066e-05,
-      "loss": 1.8676,
-      "step": 606000
-    },
-    {
-      "epoch": 8.73,
-      "learning_rate": 3.805160479595387e-05,
-      "loss": 1.8725,
-      "step": 606500
-    },
-    {
-      "epoch": 8.74,
-      "learning_rate": 3.799967805420113e-05,
-      "loss": 1.8712,
-      "step": 607000
-    },
-    {
-      "epoch": 8.74,
-      "learning_rate": 3.7947855165931904e-05,
-      "loss": 1.8725,
-      "step": 607500
-    },
-    {
-      "epoch": 8.75,
-      "learning_rate": 3.789592842417917e-05,
-      "loss": 1.8686,
-      "step": 608000
-    },
-    {
-      "epoch": 8.76,
-      "learning_rate": 3.7844001682426434e-05,
-      "loss": 1.8711,
-      "step": 608500
-    },
-    {
-      "epoch": 8.77,
-      "learning_rate": 3.77920749406737e-05,
-      "loss": 1.8691,
-      "step": 609000
-    },
-    {
-      "epoch": 8.77,
-      "learning_rate": 3.7740148198920964e-05,
-      "loss": 1.8749,
-      "step": 609500
-    },
-    {
-      "epoch": 8.78,
-      "learning_rate": 3.7688221457168225e-05,
-      "loss": 1.868,
-      "step": 610000
-    },
-    {
-      "epoch": 8.79,
-      "learning_rate": 3.7636398568899e-05,
-      "loss": 1.8713,
-      "step": 610500
-    },
-    {
-      "epoch": 8.79,
-      "learning_rate": 3.758447182714626e-05,
-      "loss": 1.8693,
-      "step": 611000
-    },
-    {
-      "epoch": 8.8,
-      "learning_rate": 3.753254508539353e-05,
-      "loss": 1.8725,
-      "step": 611500
-    },
-    {
-      "epoch": 8.81,
-      "learning_rate": 3.7480618343640797e-05,
-      "loss": 1.8677,
-      "step": 612000
-    },
-    {
-      "epoch": 8.82,
-      "learning_rate": 3.742869160188806e-05,
-      "loss": 1.8645,
-      "step": 612500
-    },
-    {
-      "epoch": 8.82,
-      "learning_rate": 3.737676486013532e-05,
-      "loss": 1.8695,
-      "step": 613000
-    },
-    {
-      "epoch": 8.83,
-      "learning_rate": 3.732494197186609e-05,
-      "loss": 1.8666,
-      "step": 613500
-    },
-    {
-      "epoch": 8.84,
-      "learning_rate": 3.7273015230113354e-05,
-      "loss": 1.8708,
-      "step": 614000
-    },
-    {
-      "epoch": 8.85,
-      "learning_rate": 3.722108848836062e-05,
-      "loss": 1.8687,
-      "step": 614500
-    },
-    {
-      "epoch": 8.85,
-      "learning_rate": 3.716916174660789e-05,
-      "loss": 1.8698,
-      "step": 615000
-    },
-    {
-      "epoch": 8.86,
-      "learning_rate": 3.711723500485515e-05,
-      "loss": 1.867,
-      "step": 615500
-    },
-    {
-      "epoch": 8.87,
-      "learning_rate": 3.7065308263102414e-05,
-      "loss": 1.8668,
-      "step": 616000
-    },
-    {
-      "epoch": 8.87,
-      "learning_rate": 3.701348537483319e-05,
-      "loss": 1.8673,
-      "step": 616500
-    },
-    {
-      "epoch": 8.88,
-      "learning_rate": 3.696155863308045e-05,
-      "loss": 1.8699,
-      "step": 617000
-    },
-    {
-      "epoch": 8.89,
-      "learning_rate": 3.6909631891327717e-05,
-      "loss": 1.8672,
-      "step": 617500
-    },
-    {
-      "epoch": 8.9,
-      "learning_rate": 3.6857705149574985e-05,
-      "loss": 1.8691,
-      "step": 618000
-    },
-    {
-      "epoch": 8.9,
-      "learning_rate": 3.6805778407822246e-05,
-      "loss": 1.8682,
-      "step": 618500
-    },
-    {
-      "epoch": 8.91,
-      "learning_rate": 3.675385166606951e-05,
-      "loss": 1.8649,
-      "step": 619000
-    },
-    {
-      "epoch": 8.92,
-      "learning_rate": 3.6701924924316776e-05,
-      "loss": 1.8702,
-      "step": 619500
-    },
-    {
-      "epoch": 8.92,
-      "learning_rate": 3.665010203604754e-05,
-      "loss": 1.8682,
-      "step": 620000
-    },
-    {
-      "epoch": 8.93,
-      "learning_rate": 3.659817529429481e-05,
-      "loss": 1.874,
-      "step": 620500
-    },
-    {
-      "epoch": 8.94,
-      "learning_rate": 3.654624855254207e-05,
-      "loss": 1.8637,
-      "step": 621000
-    },
-    {
-      "epoch": 8.95,
-      "learning_rate": 3.649432181078934e-05,
-      "loss": 1.8706,
-      "step": 621500
-    },
-    {
-      "epoch": 8.95,
-      "learning_rate": 3.644239506903661e-05,
-      "loss": 1.8685,
-      "step": 622000
-    },
-    {
-      "epoch": 8.96,
-      "learning_rate": 3.6390572180767375e-05,
-      "loss": 1.8641,
-      "step": 622500
-    },
-    {
-      "epoch": 8.97,
-      "learning_rate": 3.633864543901464e-05,
-      "loss": 1.8654,
-      "step": 623000
-    },
-    {
-      "epoch": 8.97,
-      "learning_rate": 3.6286718697261905e-05,
-      "loss": 1.8722,
-      "step": 623500
-    },
-    {
-      "epoch": 8.98,
-      "learning_rate": 3.6234791955509166e-05,
-      "loss": 1.8645,
-      "step": 624000
-    },
-    {
-      "epoch": 8.99,
-      "learning_rate": 3.6182865213756435e-05,
-      "loss": 1.8667,
-      "step": 624500
-    },
-    {
-      "epoch": 9.0,
-      "learning_rate": 3.61309384720037e-05,
-      "loss": 1.8684,
-      "step": 625000
-    },
-    {
-      "epoch": 9.0,
-      "eval_accuracy": 0.6439520651638971,
-      "eval_loss": 1.7282428741455078,
-      "eval_runtime": 647.4593,
-      "eval_samples_per_second": 832.387,
-      "eval_steps_per_second": 34.683,
-      "step": 625257
-    },
-    {
-      "epoch": 9.0,
-      "learning_rate": 3.6079011730250964e-05,
-      "loss": 1.8635,
-      "step": 625500
-    },
-    {
-      "epoch": 9.01,
-      "learning_rate": 3.6027084988498226e-05,
-      "loss": 1.8644,
-      "step": 626000
-    },
-    {
-      "epoch": 9.02,
-      "learning_rate": 3.5975262100229e-05,
-      "loss": 1.8647,
-      "step": 626500
-    },
-    {
-      "epoch": 9.03,
-      "learning_rate": 3.592333535847626e-05,
-      "loss": 1.8593,
-      "step": 627000
-    },
-    {
-      "epoch": 9.03,
-      "learning_rate": 3.587140861672353e-05,
-      "loss": 1.8659,
-      "step": 627500
-    },
-    {
-      "epoch": 9.04,
-      "learning_rate": 3.58194818749708e-05,
-      "loss": 1.8667,
-      "step": 628000
-    },
-    {
-      "epoch": 9.05,
-      "learning_rate": 3.576755513321806e-05,
-      "loss": 1.8628,
-      "step": 628500
-    },
-    {
-      "epoch": 9.05,
-      "learning_rate": 3.571573224494883e-05,
-      "loss": 1.8635,
-      "step": 629000
-    },
-    {
-      "epoch": 9.06,
-      "learning_rate": 3.566380550319609e-05,
-      "loss": 1.8626,
-      "step": 629500
-    },
-    {
-      "epoch": 9.07,
-      "learning_rate": 3.5611878761443355e-05,
-      "loss": 1.8648,
-      "step": 630000
-    },
-    {
-      "epoch": 9.08,
-      "learning_rate": 3.555995201969062e-05,
-      "loss": 1.8633,
-      "step": 630500
-    },
-    {
-      "epoch": 9.08,
-      "learning_rate": 3.550812913142139e-05,
-      "loss": 1.8675,
-      "step": 631000
-    },
-    {
-      "epoch": 9.09,
-      "learning_rate": 3.545620238966866e-05,
-      "loss": 1.8628,
-      "step": 631500
-    },
-    {
-      "epoch": 9.1,
-      "learning_rate": 3.5404275647915926e-05,
-      "loss": 1.8607,
-      "step": 632000
-    },
-    {
-      "epoch": 9.1,
-      "learning_rate": 3.535234890616319e-05,
-      "loss": 1.8607,
-      "step": 632500
-    },
-    {
-      "epoch": 9.11,
-      "learning_rate": 3.530052601789396e-05,
-      "loss": 1.8634,
-      "step": 633000
-    },
-    {
-      "epoch": 9.12,
-      "learning_rate": 3.524859927614122e-05,
-      "loss": 1.8668,
-      "step": 633500
-    },
-    {
-      "epoch": 9.13,
-      "learning_rate": 3.519667253438848e-05,
-      "loss": 1.8627,
-      "step": 634000
-    },
-    {
-      "epoch": 9.13,
-      "learning_rate": 3.514474579263575e-05,
-      "loss": 1.8585,
-      "step": 634500
-    },
-    {
-      "epoch": 9.14,
-      "learning_rate": 3.509281905088301e-05,
-      "loss": 1.8609,
-      "step": 635000
-    },
-    {
-      "epoch": 9.15,
-      "learning_rate": 3.504089230913028e-05,
-      "loss": 1.8661,
-      "step": 635500
-    },
-    {
-      "epoch": 9.15,
-      "learning_rate": 3.498896556737754e-05,
-      "loss": 1.8604,
-      "step": 636000
-    },
-    {
-      "epoch": 9.16,
-      "learning_rate": 3.493703882562481e-05,
-      "loss": 1.8576,
-      "step": 636500
-    },
-    {
-      "epoch": 9.17,
-      "learning_rate": 3.488531979083909e-05,
-      "loss": 1.8629,
-      "step": 637000
-    },
-    {
-      "epoch": 9.18,
-      "learning_rate": 3.483339304908635e-05,
-      "loss": 1.8623,
-      "step": 637500
-    },
-    {
-      "epoch": 9.18,
-      "learning_rate": 3.478146630733362e-05,
-      "loss": 1.8646,
-      "step": 638000
-    },
-    {
-      "epoch": 9.19,
-      "learning_rate": 3.472953956558088e-05,
-      "loss": 1.8596,
-      "step": 638500
-    },
-    {
-      "epoch": 9.2,
-      "learning_rate": 3.467761282382814e-05,
-      "loss": 1.8588,
-      "step": 639000
-    },
-    {
-      "epoch": 9.21,
-      "learning_rate": 3.462568608207541e-05,
-      "loss": 1.8619,
-      "step": 639500
-    },
-    {
-      "epoch": 9.21,
-      "learning_rate": 3.457375934032268e-05,
-      "loss": 1.8561,
-      "step": 640000
-    },
-    {
-      "epoch": 9.22,
-      "learning_rate": 3.452183259856993e-05,
-      "loss": 1.8604,
-      "step": 640500
-    },
-    {
-      "epoch": 9.23,
-      "learning_rate": 3.447000971030071e-05,
-      "loss": 1.8586,
-      "step": 641000
-    },
-    {
-      "epoch": 9.23,
-      "learning_rate": 3.4418082968547974e-05,
-      "loss": 1.8609,
-      "step": 641500
-    },
-    {
-      "epoch": 9.24,
-      "learning_rate": 3.4366156226795236e-05,
-      "loss": 1.8588,
-      "step": 642000
-    },
-    {
-      "epoch": 9.25,
-      "learning_rate": 3.431433333852601e-05,
-      "loss": 1.8609,
-      "step": 642500
-    },
-    {
-      "epoch": 9.26,
-      "learning_rate": 3.426240659677327e-05,
-      "loss": 1.8614,
-      "step": 643000
-    },
-    {
-      "epoch": 9.26,
-      "learning_rate": 3.421047985502054e-05,
-      "loss": 1.8554,
-      "step": 643500
-    },
-    {
-      "epoch": 9.27,
-      "learning_rate": 3.415855311326781e-05,
-      "loss": 1.8618,
-      "step": 644000
-    },
-    {
-      "epoch": 9.28,
-      "learning_rate": 3.410662637151507e-05,
-      "loss": 1.8614,
-      "step": 644500
-    },
-    {
-      "epoch": 9.28,
-      "learning_rate": 3.405480348324584e-05,
-      "loss": 1.8586,
-      "step": 645000
-    },
-    {
-      "epoch": 9.29,
-      "learning_rate": 3.40028767414931e-05,
-      "loss": 1.8631,
-      "step": 645500
-    },
-    {
-      "epoch": 9.3,
-      "learning_rate": 3.3950949999740365e-05,
-      "loss": 1.8605,
-      "step": 646000
-    },
-    {
-      "epoch": 9.31,
-      "learning_rate": 3.389902325798763e-05,
-      "loss": 1.862,
-      "step": 646500
-    },
-    {
-      "epoch": 9.31,
-      "learning_rate": 3.38472003697184e-05,
-      "loss": 1.8579,
-      "step": 647000
-    },
-    {
-      "epoch": 9.32,
-      "learning_rate": 3.379527362796567e-05,
-      "loss": 1.8596,
-      "step": 647500
-    },
-    {
-      "epoch": 9.33,
-      "learning_rate": 3.3743346886212936e-05,
-      "loss": 1.8614,
-      "step": 648000
-    },
-    {
-      "epoch": 9.33,
-      "learning_rate": 3.36914201444602e-05,
-      "loss": 1.862,
-      "step": 648500
-    },
-    {
-      "epoch": 9.34,
-      "learning_rate": 3.363949340270746e-05,
-      "loss": 1.8548,
-      "step": 649000
-    },
-    {
-      "epoch": 9.35,
-      "learning_rate": 3.358767051443823e-05,
-      "loss": 1.861,
-      "step": 649500
-    },
-    {
-      "epoch": 9.36,
-      "learning_rate": 3.353574377268549e-05,
-      "loss": 1.8625,
-      "step": 650000
-    },
-    {
-      "epoch": 9.36,
-      "learning_rate": 3.348381703093276e-05,
-      "loss": 1.8584,
-      "step": 650500
-    },
-    {
-      "epoch": 9.37,
-      "learning_rate": 3.343189028918003e-05,
-      "loss": 1.859,
-      "step": 651000
-    },
-    {
-      "epoch": 9.38,
-      "learning_rate": 3.3380067400910796e-05,
-      "loss": 1.8593,
-      "step": 651500
-    },
-    {
-      "epoch": 9.38,
-      "learning_rate": 3.3328140659158064e-05,
-      "loss": 1.86,
-      "step": 652000
-    },
-    {
-      "epoch": 9.39,
-      "learning_rate": 3.3276213917405326e-05,
-      "loss": 1.8634,
-      "step": 652500
-    },
-    {
-      "epoch": 9.4,
-      "learning_rate": 3.322428717565259e-05,
-      "loss": 1.8623,
-      "step": 653000
-    },
-    {
-      "epoch": 9.41,
-      "learning_rate": 3.3172360433899856e-05,
-      "loss": 1.8594,
-      "step": 653500
-    },
-    {
-      "epoch": 9.41,
-      "learning_rate": 3.312043369214712e-05,
-      "loss": 1.8619,
-      "step": 654000
-    },
-    {
-      "epoch": 9.42,
-      "learning_rate": 3.306861080387789e-05,
-      "loss": 1.8597,
-      "step": 654500
-    },
-    {
-      "epoch": 9.43,
-      "learning_rate": 3.301668406212515e-05,
-      "loss": 1.8595,
-      "step": 655000
-    },
-    {
-      "epoch": 9.44,
-      "learning_rate": 3.296475732037242e-05,
-      "loss": 1.8574,
-      "step": 655500
-    },
-    {
-      "epoch": 9.44,
-      "learning_rate": 3.291283057861969e-05,
-      "loss": 1.8601,
-      "step": 656000
-    },
-    {
-      "epoch": 9.45,
-      "learning_rate": 3.286090383686695e-05,
-      "loss": 1.8594,
-      "step": 656500
-    },
-    {
-      "epoch": 9.46,
-      "learning_rate": 3.280908094859772e-05,
-      "loss": 1.8607,
-      "step": 657000
-    },
-    {
-      "epoch": 9.46,
-      "learning_rate": 3.2757154206844984e-05,
-      "loss": 1.8589,
-      "step": 657500
-    },
-    {
-      "epoch": 9.47,
-      "learning_rate": 3.2705227465092246e-05,
-      "loss": 1.8601,
-      "step": 658000
-    },
-    {
-      "epoch": 9.48,
-      "learning_rate": 3.2653300723339514e-05,
-      "loss": 1.8606,
-      "step": 658500
-    },
-    {
-      "epoch": 9.49,
-      "learning_rate": 3.260137398158678e-05,
-      "loss": 1.8601,
-      "step": 659000
-    },
-    {
-      "epoch": 9.49,
-      "learning_rate": 3.2549447239834044e-05,
-      "loss": 1.8591,
-      "step": 659500
-    },
-    {
-      "epoch": 9.5,
-      "learning_rate": 3.249762435156482e-05,
-      "loss": 1.8624,
-      "step": 660000
-    },
-    {
-      "epoch": 9.51,
-      "learning_rate": 3.244569760981208e-05,
-      "loss": 1.8576,
-      "step": 660500
-    },
-    {
-      "epoch": 9.51,
-      "learning_rate": 3.239377086805934e-05,
-      "loss": 1.8572,
-      "step": 661000
-    },
-    {
-      "epoch": 9.52,
-      "learning_rate": 3.234184412630661e-05,
-      "loss": 1.8566,
-      "step": 661500
-    },
-    {
-      "epoch": 9.53,
-      "learning_rate": 3.228991738455388e-05,
-      "loss": 1.861,
-      "step": 662000
-    },
-    {
-      "epoch": 9.54,
-      "learning_rate": 3.223809449628464e-05,
-      "loss": 1.8551,
-      "step": 662500
-    },
-    {
-      "epoch": 9.54,
-      "learning_rate": 3.218616775453191e-05,
-      "loss": 1.8572,
-      "step": 663000
-    },
-    {
-      "epoch": 9.55,
-      "learning_rate": 3.213424101277917e-05,
-      "loss": 1.8547,
-      "step": 663500
-    },
-    {
-      "epoch": 9.56,
-      "learning_rate": 3.2082314271026434e-05,
-      "loss": 1.8563,
-      "step": 664000
-    },
-    {
-      "epoch": 9.56,
-      "learning_rate": 3.20303875292737e-05,
-      "loss": 1.8589,
-      "step": 664500
-    },
-    {
-      "epoch": 9.57,
-      "learning_rate": 3.197846078752097e-05,
-      "loss": 1.8556,
-      "step": 665000
-    },
-    {
-      "epoch": 9.58,
-      "learning_rate": 3.192653404576823e-05,
-      "loss": 1.8538,
-      "step": 665500
-    },
-    {
-      "epoch": 9.59,
-      "learning_rate": 3.1874711157499005e-05,
-      "loss": 1.8587,
-      "step": 666000
-    },
-    {
-      "epoch": 9.59,
-      "learning_rate": 3.182278441574627e-05,
-      "loss": 1.8566,
-      "step": 666500
-    },
-    {
-      "epoch": 9.6,
-      "learning_rate": 3.177085767399353e-05,
-      "loss": 1.8564,
-      "step": 667000
-    },
-    {
-      "epoch": 9.61,
-      "learning_rate": 3.17189309322408e-05,
-      "loss": 1.8565,
-      "step": 667500
-    },
-    {
-      "epoch": 9.62,
-      "learning_rate": 3.166700419048806e-05,
-      "loss": 1.8574,
-      "step": 668000
-    },
-    {
-      "epoch": 9.62,
-      "learning_rate": 3.161518130221883e-05,
-      "loss": 1.8576,
-      "step": 668500
-    },
-    {
-      "epoch": 9.63,
-      "learning_rate": 3.156325456046609e-05,
-      "loss": 1.8628,
-      "step": 669000
-    },
-    {
-      "epoch": 9.64,
-      "learning_rate": 3.151132781871336e-05,
-      "loss": 1.8601,
-      "step": 669500
-    },
-    {
-      "epoch": 9.64,
-      "learning_rate": 3.145940107696063e-05,
-      "loss": 1.8549,
-      "step": 670000
-    },
-    {
-      "epoch": 9.65,
-      "learning_rate": 3.140747433520789e-05,
-      "loss": 1.8548,
-      "step": 670500
-    },
-    {
-      "epoch": 9.66,
-      "learning_rate": 3.135554759345515e-05,
-      "loss": 1.8594,
-      "step": 671000
-    },
-    {
-      "epoch": 9.67,
-      "learning_rate": 3.130362085170242e-05,
-      "loss": 1.8564,
-      "step": 671500
-    },
-    {
-      "epoch": 9.67,
-      "learning_rate": 3.125169410994969e-05,
-      "loss": 1.8547,
-      "step": 672000
-    },
-    {
-      "epoch": 9.68,
-      "learning_rate": 3.1199871221680455e-05,
-      "loss": 1.8538,
-      "step": 672500
-    },
-    {
-      "epoch": 9.69,
-      "learning_rate": 3.114804833341122e-05,
-      "loss": 1.8571,
-      "step": 673000
-    },
-    {
-      "epoch": 9.69,
-      "learning_rate": 3.109612159165849e-05,
-      "loss": 1.855,
-      "step": 673500
-    },
-    {
-      "epoch": 9.7,
-      "learning_rate": 3.104419484990576e-05,
-      "loss": 1.8553,
-      "step": 674000
-    },
-    {
-      "epoch": 9.71,
-      "learning_rate": 3.099226810815302e-05,
-      "loss": 1.8581,
-      "step": 674500
-    },
-    {
-      "epoch": 9.72,
-      "learning_rate": 3.094034136640028e-05,
-      "loss": 1.8542,
-      "step": 675000
-    },
-    {
-      "epoch": 9.72,
-      "learning_rate": 3.088841462464755e-05,
-      "loss": 1.8521,
-      "step": 675500
-    },
-    {
-      "epoch": 9.73,
-      "learning_rate": 3.083648788289482e-05,
-      "loss": 1.8558,
-      "step": 676000
-    },
-    {
-      "epoch": 9.74,
-      "learning_rate": 3.0784664994625584e-05,
-      "loss": 1.8517,
-      "step": 676500
-    },
-    {
-      "epoch": 9.74,
-      "learning_rate": 3.073273825287285e-05,
-      "loss": 1.8526,
-      "step": 677000
-    },
-    {
-      "epoch": 9.75,
-      "learning_rate": 3.0680811511120114e-05,
-      "loss": 1.8609,
-      "step": 677500
-    },
-    {
-      "epoch": 9.76,
-      "learning_rate": 3.0628884769367375e-05,
-      "loss": 1.853,
-      "step": 678000
-    },
-    {
-      "epoch": 9.77,
-      "learning_rate": 3.0576958027614643e-05,
-      "loss": 1.8541,
-      "step": 678500
-    },
-    {
-      "epoch": 9.77,
-      "learning_rate": 3.052513513934541e-05,
-      "loss": 1.8562,
-      "step": 679000
-    },
-    {
-      "epoch": 9.78,
-      "learning_rate": 3.0473208397592678e-05,
-      "loss": 1.852,
-      "step": 679500
-    },
-    {
-      "epoch": 9.79,
-      "learning_rate": 3.0421281655839946e-05,
-      "loss": 1.8511,
-      "step": 680000
-    },
-    {
-      "epoch": 9.8,
-      "learning_rate": 3.0369354914087204e-05,
-      "loss": 1.8544,
-      "step": 680500
-    },
-    {
-      "epoch": 9.8,
-      "learning_rate": 3.031753202581798e-05,
-      "loss": 1.8553,
-      "step": 681000
-    },
-    {
-      "epoch": 9.81,
-      "learning_rate": 3.026560528406524e-05,
-      "loss": 1.859,
-      "step": 681500
-    },
-    {
-      "epoch": 9.82,
-      "learning_rate": 3.0213678542312507e-05,
-      "loss": 1.8608,
-      "step": 682000
-    },
-    {
-      "epoch": 9.82,
-      "learning_rate": 3.0161751800559772e-05,
-      "loss": 1.8512,
-      "step": 682500
-    },
-    {
-      "epoch": 9.83,
-      "learning_rate": 3.0109928912290542e-05,
-      "loss": 1.8558,
-      "step": 683000
-    },
-    {
-      "epoch": 9.84,
-      "learning_rate": 3.0058002170537807e-05,
-      "loss": 1.8516,
-      "step": 683500
-    },
-    {
-      "epoch": 9.85,
-      "learning_rate": 3.0006075428785068e-05,
-      "loss": 1.8533,
-      "step": 684000
-    },
-    {
-      "epoch": 9.85,
-      "learning_rate": 2.9954148687032337e-05,
-      "loss": 1.856,
-      "step": 684500
-    },
-    {
-      "epoch": 9.86,
-      "learning_rate": 2.99022219452796e-05,
-      "loss": 1.8492,
-      "step": 685000
-    },
-    {
-      "epoch": 9.87,
-      "learning_rate": 2.985039905701037e-05,
-      "loss": 1.8532,
-      "step": 685500
-    },
-    {
-      "epoch": 9.87,
-      "learning_rate": 2.9798472315257636e-05,
-      "loss": 1.856,
-      "step": 686000
-    },
-    {
-      "epoch": 9.88,
-      "learning_rate": 2.9746545573504904e-05,
-      "loss": 1.8571,
-      "step": 686500
-    },
-    {
-      "epoch": 9.89,
-      "learning_rate": 2.9694618831752162e-05,
-      "loss": 1.8554,
-      "step": 687000
-    },
-    {
-      "epoch": 9.9,
-      "learning_rate": 2.964269208999943e-05,
-      "loss": 1.8535,
-      "step": 687500
-    },
-    {
-      "epoch": 9.9,
-      "learning_rate": 2.9590765348246696e-05,
-      "loss": 1.853,
-      "step": 688000
-    },
-    {
-      "epoch": 9.91,
-      "learning_rate": 2.9538838606493957e-05,
-      "loss": 1.8502,
-      "step": 688500
-    },
-    {
-      "epoch": 9.92,
-      "learning_rate": 2.948701571822473e-05,
-      "loss": 1.8495,
-      "step": 689000
-    },
-    {
-      "epoch": 9.92,
-      "learning_rate": 2.9435088976472e-05,
-      "loss": 1.8513,
-      "step": 689500
-    },
-    {
-      "epoch": 9.93,
-      "learning_rate": 2.9383162234719257e-05,
-      "loss": 1.8521,
-      "step": 690000
-    },
-    {
-      "epoch": 9.94,
-      "learning_rate": 2.9331235492966525e-05,
-      "loss": 1.8537,
-      "step": 690500
-    },
-    {
-      "epoch": 9.95,
-      "learning_rate": 2.927930875121379e-05,
-      "loss": 1.8503,
-      "step": 691000
-    },
-    {
-      "epoch": 9.95,
-      "learning_rate": 2.922738200946105e-05,
-      "loss": 1.8505,
-      "step": 691500
-    },
-    {
-      "epoch": 9.96,
-      "learning_rate": 2.9175559121191824e-05,
-      "loss": 1.8586,
-      "step": 692000
-    },
-    {
-      "epoch": 9.97,
-      "learning_rate": 2.9123632379439086e-05,
-      "loss": 1.8524,
-      "step": 692500
-    },
-    {
-      "epoch": 9.98,
-      "learning_rate": 2.9071705637686354e-05,
-      "loss": 1.8504,
-      "step": 693000
-    },
-    {
-      "epoch": 9.98,
-      "learning_rate": 2.901977889593362e-05,
-      "loss": 1.8514,
-      "step": 693500
-    },
-    {
-      "epoch": 9.99,
-      "learning_rate": 2.8967852154180887e-05,
-      "loss": 1.8511,
-      "step": 694000
-    },
-    {
-      "epoch": 10.0,
-      "learning_rate": 2.8915925412428145e-05,
-      "loss": 1.8517,
-      "step": 694500
-    },
-    {
-      "epoch": 10.0,
-      "eval_accuracy": 0.6463801918157677,
-      "eval_loss": 1.7140141725540161,
-      "eval_runtime": 646.6238,
-      "eval_samples_per_second": 833.463,
-      "eval_steps_per_second": 34.728,
-      "step": 694730
-    },
-    {
-      "epoch": 10.0,
-      "learning_rate": 2.8863998670675414e-05,
-      "loss": 1.851,
-      "step": 695000
-    },
-    {
-      "epoch": 10.01,
-      "learning_rate": 2.881217578240618e-05,
-      "loss": 1.8505,
-      "step": 695500
-    },
-    {
-      "epoch": 10.02,
-      "learning_rate": 2.8760249040653448e-05,
-      "loss": 1.8464,
-      "step": 696000
-    },
-    {
-      "epoch": 10.03,
-      "learning_rate": 2.8708322298900713e-05,
-      "loss": 1.852,
-      "step": 696500
-    },
-    {
-      "epoch": 10.03,
-      "learning_rate": 2.8656395557147975e-05,
-      "loss": 1.8485,
-      "step": 697000
-    },
-    {
-      "epoch": 10.04,
-      "learning_rate": 2.860446881539524e-05,
-      "loss": 1.8525,
-      "step": 697500
-    },
-    {
-      "epoch": 10.05,
-      "learning_rate": 2.855264592712601e-05,
-      "loss": 1.851,
-      "step": 698000
-    },
-    {
-      "epoch": 10.05,
-      "learning_rate": 2.8500719185373277e-05,
-      "loss": 1.8502,
-      "step": 698500
-    },
-    {
-      "epoch": 10.06,
-      "learning_rate": 2.8448792443620542e-05,
-      "loss": 1.8492,
-      "step": 699000
-    },
-    {
-      "epoch": 10.07,
-      "learning_rate": 2.8396865701867807e-05,
-      "loss": 1.849,
-      "step": 699500
-    },
-    {
-      "epoch": 10.08,
-      "learning_rate": 2.834493896011507e-05,
-      "loss": 1.8477,
-      "step": 700000
-    },
-    {
-      "epoch": 10.08,
-      "learning_rate": 2.8293012218362337e-05,
-      "loss": 1.8453,
-      "step": 700500
-    },
-    {
-      "epoch": 10.09,
-      "learning_rate": 2.8241189330093103e-05,
-      "loss": 1.8462,
-      "step": 701000
-    },
-    {
-      "epoch": 10.1,
-      "learning_rate": 2.818926258834037e-05,
-      "loss": 1.8501,
-      "step": 701500
-    },
-    {
-      "epoch": 10.1,
-      "learning_rate": 2.8137335846587636e-05,
-      "loss": 1.8469,
-      "step": 702000
-    },
-    {
-      "epoch": 10.11,
-      "learning_rate": 2.8085409104834898e-05,
-      "loss": 1.8491,
-      "step": 702500
-    },
-    {
-      "epoch": 10.12,
-      "learning_rate": 2.8033482363082163e-05,
-      "loss": 1.8436,
-      "step": 703000
-    },
-    {
-      "epoch": 10.13,
-      "learning_rate": 2.798155562132943e-05,
-      "loss": 1.851,
-      "step": 703500
-    },
-    {
-      "epoch": 10.13,
-      "learning_rate": 2.7929628879576696e-05,
-      "loss": 1.8514,
-      "step": 704000
-    },
-    {
-      "epoch": 10.14,
-      "learning_rate": 2.7877702137823958e-05,
-      "loss": 1.8483,
-      "step": 704500
-    },
-    {
-      "epoch": 10.15,
-      "learning_rate": 2.782587924955473e-05,
-      "loss": 1.8516,
-      "step": 705000
-    },
-    {
-      "epoch": 10.16,
-      "learning_rate": 2.7773952507801992e-05,
-      "loss": 1.8502,
-      "step": 705500
-    },
-    {
-      "epoch": 10.16,
-      "learning_rate": 2.772202576604926e-05,
-      "loss": 1.8506,
-      "step": 706000
-    },
-    {
-      "epoch": 10.17,
-      "learning_rate": 2.7670099024296525e-05,
-      "loss": 1.8464,
-      "step": 706500
-    },
-    {
-      "epoch": 10.18,
-      "learning_rate": 2.76183799895108e-05,
-      "loss": 1.8482,
-      "step": 707000
-    },
-    {
-      "epoch": 10.18,
-      "learning_rate": 2.756645324775806e-05,
-      "loss": 1.8451,
-      "step": 707500
-    },
-    {
-      "epoch": 10.19,
-      "learning_rate": 2.751452650600533e-05,
-      "loss": 1.851,
-      "step": 708000
-    },
-    {
-      "epoch": 10.2,
-      "learning_rate": 2.7462599764252594e-05,
-      "loss": 1.8453,
-      "step": 708500
-    },
-    {
-      "epoch": 10.21,
-      "learning_rate": 2.741067302249986e-05,
-      "loss": 1.8466,
-      "step": 709000
-    },
-    {
-      "epoch": 10.21,
-      "learning_rate": 2.735885013423063e-05,
-      "loss": 1.8497,
-      "step": 709500
-    },
-    {
-      "epoch": 10.22,
-      "learning_rate": 2.7306923392477897e-05,
-      "loss": 1.8495,
-      "step": 710000
-    },
-    {
-      "epoch": 10.23,
-      "learning_rate": 2.7254996650725155e-05,
-      "loss": 1.8506,
-      "step": 710500
-    },
-    {
-      "epoch": 10.23,
-      "learning_rate": 2.7203069908972424e-05,
-      "loss": 1.8451,
-      "step": 711000
-    },
-    {
-      "epoch": 10.24,
-      "learning_rate": 2.715114316721969e-05,
-      "loss": 1.8486,
-      "step": 711500
-    },
-    {
-      "epoch": 10.25,
-      "learning_rate": 2.709921642546695e-05,
-      "loss": 1.8487,
-      "step": 712000
-    },
-    {
-      "epoch": 10.26,
-      "learning_rate": 2.7047289683714215e-05,
-      "loss": 1.8479,
-      "step": 712500
-    },
-    {
-      "epoch": 10.26,
-      "learning_rate": 2.699546679544499e-05,
-      "loss": 1.8466,
-      "step": 713000
-    },
-    {
-      "epoch": 10.27,
-      "learning_rate": 2.694354005369225e-05,
-      "loss": 1.8496,
-      "step": 713500
-    },
-    {
-      "epoch": 10.28,
-      "learning_rate": 2.6891613311939518e-05,
-      "loss": 1.8452,
-      "step": 714000
-    },
-    {
-      "epoch": 10.28,
-      "learning_rate": 2.6839686570186783e-05,
-      "loss": 1.8497,
-      "step": 714500
-    },
-    {
-      "epoch": 10.29,
-      "learning_rate": 2.6787759828434044e-05,
-      "loss": 1.8442,
-      "step": 715000
-    },
-    {
-      "epoch": 10.3,
-      "learning_rate": 2.6735936940164817e-05,
-      "loss": 1.8458,
-      "step": 715500
-    },
-    {
-      "epoch": 10.31,
-      "learning_rate": 2.668401019841208e-05,
-      "loss": 1.8462,
-      "step": 716000
-    },
-    {
-      "epoch": 10.31,
-      "learning_rate": 2.6632083456659347e-05,
-      "loss": 1.846,
-      "step": 716500
-    },
-    {
-      "epoch": 10.32,
-      "learning_rate": 2.6580156714906612e-05,
-      "loss": 1.8428,
-      "step": 717000
-    },
-    {
-      "epoch": 10.33,
-      "learning_rate": 2.652833382663738e-05,
-      "loss": 1.8453,
-      "step": 717500
-    },
-    {
-      "epoch": 10.33,
-      "learning_rate": 2.6476407084884646e-05,
-      "loss": 1.8468,
-      "step": 718000
-    },
-    {
-      "epoch": 10.34,
-      "learning_rate": 2.6424480343131915e-05,
-      "loss": 1.8439,
-      "step": 718500
-    },
-    {
-      "epoch": 10.35,
-      "learning_rate": 2.6372553601379173e-05,
-      "loss": 1.8415,
-      "step": 719000
-    },
-    {
-      "epoch": 10.36,
-      "learning_rate": 2.632062685962644e-05,
-      "loss": 1.8476,
-      "step": 719500
-    },
-    {
-      "epoch": 10.36,
-      "learning_rate": 2.6268803971357207e-05,
-      "loss": 1.8418,
-      "step": 720000
-    },
-    {
-      "epoch": 10.37,
-      "learning_rate": 2.6216877229604476e-05,
-      "loss": 1.8449,
-      "step": 720500
-    },
-    {
-      "epoch": 10.38,
-      "learning_rate": 2.616495048785174e-05,
-      "loss": 1.8443,
-      "step": 721000
-    },
-    {
-      "epoch": 10.39,
-      "learning_rate": 2.6113023746099002e-05,
-      "loss": 1.8463,
-      "step": 721500
-    },
-    {
-      "epoch": 10.39,
-      "learning_rate": 2.606109700434627e-05,
-      "loss": 1.8501,
-      "step": 722000
-    },
-    {
-      "epoch": 10.4,
-      "learning_rate": 2.6009170262593535e-05,
-      "loss": 1.8479,
-      "step": 722500
-    },
-    {
-      "epoch": 10.41,
-      "learning_rate": 2.5957347374324305e-05,
-      "loss": 1.8497,
-      "step": 723000
-    },
-    {
-      "epoch": 10.41,
-      "learning_rate": 2.590542063257157e-05,
-      "loss": 1.8438,
-      "step": 723500
-    },
-    {
-      "epoch": 10.42,
-      "learning_rate": 2.5853493890818835e-05,
-      "loss": 1.8444,
-      "step": 724000
-    },
-    {
-      "epoch": 10.43,
-      "learning_rate": 2.5801567149066096e-05,
-      "loss": 1.8485,
-      "step": 724500
-    },
-    {
-      "epoch": 10.44,
-      "learning_rate": 2.5749640407313365e-05,
-      "loss": 1.8456,
-      "step": 725000
-    },
-    {
-      "epoch": 10.44,
-      "learning_rate": 2.569781751904413e-05,
-      "loss": 1.846,
-      "step": 725500
-    },
-    {
-      "epoch": 10.45,
-      "learning_rate": 2.56458907772914e-05,
-      "loss": 1.8492,
-      "step": 726000
-    },
-    {
-      "epoch": 10.46,
-      "learning_rate": 2.5593964035538664e-05,
-      "loss": 1.8509,
-      "step": 726500
-    },
-    {
-      "epoch": 10.46,
-      "learning_rate": 2.5542037293785926e-05,
-      "loss": 1.8415,
-      "step": 727000
-    },
-    {
-      "epoch": 10.47,
-      "learning_rate": 2.549011055203319e-05,
-      "loss": 1.8446,
-      "step": 727500
-    },
-    {
-      "epoch": 10.48,
-      "learning_rate": 2.5438287663763967e-05,
-      "loss": 1.8488,
-      "step": 728000
-    },
-    {
-      "epoch": 10.49,
-      "learning_rate": 2.5386360922011225e-05,
-      "loss": 1.8439,
-      "step": 728500
-    },
-    {
-      "epoch": 10.49,
-      "learning_rate": 2.5334434180258493e-05,
-      "loss": 1.8464,
-      "step": 729000
-    },
-    {
-      "epoch": 10.5,
-      "learning_rate": 2.5282507438505758e-05,
-      "loss": 1.8486,
-      "step": 729500
-    },
-    {
-      "epoch": 10.51,
-      "learning_rate": 2.523058069675302e-05,
-      "loss": 1.8476,
-      "step": 730000
-    },
-    {
-      "epoch": 10.51,
-      "learning_rate": 2.5178653955000288e-05,
-      "loss": 1.8451,
-      "step": 730500
-    },
-    {
-      "epoch": 10.52,
-      "learning_rate": 2.5126727213247553e-05,
-      "loss": 1.8463,
-      "step": 731000
-    },
-    {
-      "epoch": 10.53,
-      "learning_rate": 2.5074904324978322e-05,
-      "loss": 1.842,
-      "step": 731500
-    },
-    {
-      "epoch": 10.54,
-      "learning_rate": 2.5022977583225587e-05,
-      "loss": 1.8431,
-      "step": 732000
-    },
-    {
-      "epoch": 10.54,
-      "learning_rate": 2.4971050841472852e-05,
-      "loss": 1.8445,
-      "step": 732500
-    },
-    {
-      "epoch": 10.55,
-      "learning_rate": 2.4919124099720114e-05,
-      "loss": 1.843,
-      "step": 733000
-    },
-    {
-      "epoch": 10.56,
-      "learning_rate": 2.4867197357967382e-05,
-      "loss": 1.8444,
-      "step": 733500
-    },
-    {
-      "epoch": 10.57,
-      "learning_rate": 2.4815374469698152e-05,
-      "loss": 1.8471,
-      "step": 734000
-    },
-    {
-      "epoch": 10.57,
-      "learning_rate": 2.4763447727945417e-05,
-      "loss": 1.8458,
-      "step": 734500
-    },
-    {
-      "epoch": 10.58,
-      "learning_rate": 2.4711520986192678e-05,
-      "loss": 1.839,
-      "step": 735000
-    },
-    {
-      "epoch": 10.59,
-      "learning_rate": 2.4659594244439946e-05,
-      "loss": 1.844,
-      "step": 735500
-    },
-    {
-      "epoch": 10.59,
-      "learning_rate": 2.4607771356170716e-05,
-      "loss": 1.8431,
-      "step": 736000
-    },
-    {
-      "epoch": 10.6,
-      "learning_rate": 2.455584461441798e-05,
-      "loss": 1.847,
-      "step": 736500
-    },
-    {
-      "epoch": 10.61,
-      "learning_rate": 2.4503917872665242e-05,
-      "loss": 1.8469,
-      "step": 737000
-    },
-    {
-      "epoch": 10.62,
-      "learning_rate": 2.445199113091251e-05,
-      "loss": 1.8411,
-      "step": 737500
-    },
-    {
-      "epoch": 10.62,
-      "learning_rate": 2.4400064389159776e-05,
-      "loss": 1.8458,
-      "step": 738000
-    },
-    {
-      "epoch": 10.63,
-      "learning_rate": 2.434813764740704e-05,
-      "loss": 1.8439,
-      "step": 738500
-    },
-    {
-      "epoch": 10.64,
-      "learning_rate": 2.429631475913781e-05,
-      "loss": 1.8413,
-      "step": 739000
-    },
-    {
-      "epoch": 10.64,
-      "learning_rate": 2.4244388017385075e-05,
-      "loss": 1.8451,
-      "step": 739500
-    },
-    {
-      "epoch": 10.65,
-      "learning_rate": 2.419246127563234e-05,
-      "loss": 1.8435,
-      "step": 740000
-    },
-    {
-      "epoch": 10.66,
-      "learning_rate": 2.4140534533879605e-05,
-      "loss": 1.8414,
-      "step": 740500
-    },
-    {
-      "epoch": 10.67,
-      "learning_rate": 2.408860779212687e-05,
-      "loss": 1.8354,
-      "step": 741000
-    },
-    {
-      "epoch": 10.67,
-      "learning_rate": 2.403678490385764e-05,
-      "loss": 1.8443,
-      "step": 741500
-    },
-    {
-      "epoch": 10.68,
-      "learning_rate": 2.3984858162104904e-05,
-      "loss": 1.8406,
-      "step": 742000
-    },
-    {
-      "epoch": 10.69,
-      "learning_rate": 2.3932931420352166e-05,
-      "loss": 1.8435,
-      "step": 742500
-    },
-    {
-      "epoch": 10.69,
-      "learning_rate": 2.3881004678599434e-05,
-      "loss": 1.8416,
-      "step": 743000
-    },
-    {
-      "epoch": 10.7,
-      "learning_rate": 2.3829181790330204e-05,
-      "loss": 1.8432,
-      "step": 743500
-    },
-    {
-      "epoch": 10.71,
-      "learning_rate": 2.377725504857747e-05,
-      "loss": 1.842,
-      "step": 744000
-    },
-    {
-      "epoch": 10.72,
-      "learning_rate": 2.372532830682473e-05,
-      "loss": 1.8392,
-      "step": 744500
-    },
-    {
-      "epoch": 10.72,
-      "learning_rate": 2.3673401565072e-05,
-      "loss": 1.8448,
-      "step": 745000
-    },
-    {
-      "epoch": 10.73,
-      "learning_rate": 2.3621474823319263e-05,
-      "loss": 1.8396,
-      "step": 745500
-    },
-    {
-      "epoch": 10.74,
-      "learning_rate": 2.3569548081566528e-05,
-      "loss": 1.8434,
-      "step": 746000
-    },
-    {
-      "epoch": 10.75,
-      "learning_rate": 2.3517621339813793e-05,
-      "loss": 1.8438,
-      "step": 746500
-    },
-    {
-      "epoch": 10.75,
-      "learning_rate": 2.3465694598061055e-05,
-      "loss": 1.8442,
-      "step": 747000
-    },
-    {
-      "epoch": 10.76,
-      "learning_rate": 2.3413975563275332e-05,
-      "loss": 1.8365,
-      "step": 747500
-    },
-    {
-      "epoch": 10.77,
-      "learning_rate": 2.3362048821522597e-05,
-      "loss": 1.8394,
-      "step": 748000
-    },
-    {
-      "epoch": 10.77,
-      "learning_rate": 2.3310122079769862e-05,
-      "loss": 1.8396,
-      "step": 748500
-    },
-    {
-      "epoch": 10.78,
-      "learning_rate": 2.3258195338017127e-05,
-      "loss": 1.8463,
-      "step": 749000
-    },
-    {
-      "epoch": 10.79,
-      "learning_rate": 2.3206268596264392e-05,
-      "loss": 1.8422,
-      "step": 749500
-    },
-    {
-      "epoch": 10.8,
-      "learning_rate": 2.3154445707995162e-05,
-      "loss": 1.8358,
-      "step": 750000
-    },
-    {
-      "epoch": 10.8,
-      "learning_rate": 2.3102518966242427e-05,
-      "loss": 1.8415,
-      "step": 750500
-    },
-    {
-      "epoch": 10.81,
-      "learning_rate": 2.305059222448969e-05,
-      "loss": 1.8384,
-      "step": 751000
-    },
-    {
-      "epoch": 10.82,
-      "learning_rate": 2.2998665482736956e-05,
-      "loss": 1.8431,
-      "step": 751500
-    },
-    {
-      "epoch": 10.82,
-      "learning_rate": 2.2946738740984218e-05,
-      "loss": 1.8381,
-      "step": 752000
-    },
-    {
-      "epoch": 10.83,
-      "learning_rate": 2.289491585271499e-05,
-      "loss": 1.8402,
-      "step": 752500
-    },
-    {
-      "epoch": 10.84,
-      "learning_rate": 2.2842989110962256e-05,
-      "loss": 1.8399,
-      "step": 753000
-    },
-    {
-      "epoch": 10.85,
-      "learning_rate": 2.279106236920952e-05,
-      "loss": 1.844,
-      "step": 753500
-    },
-    {
-      "epoch": 10.85,
-      "learning_rate": 2.2739135627456786e-05,
-      "loss": 1.8441,
-      "step": 754000
-    },
-    {
-      "epoch": 10.86,
-      "learning_rate": 2.2687312739187555e-05,
-      "loss": 1.8416,
-      "step": 754500
-    },
-    {
-      "epoch": 10.87,
-      "learning_rate": 2.263538599743482e-05,
-      "loss": 1.8406,
-      "step": 755000
-    },
-    {
-      "epoch": 10.87,
-      "learning_rate": 2.2583459255682085e-05,
-      "loss": 1.8428,
-      "step": 755500
-    },
-    {
-      "epoch": 10.88,
-      "learning_rate": 2.253153251392935e-05,
-      "loss": 1.8416,
-      "step": 756000
-    },
-    {
-      "epoch": 10.89,
-      "learning_rate": 2.2479605772176615e-05,
-      "loss": 1.8425,
-      "step": 756500
-    },
-    {
-      "epoch": 10.9,
-      "learning_rate": 2.242767903042388e-05,
-      "loss": 1.8403,
-      "step": 757000
-    },
-    {
-      "epoch": 10.9,
-      "learning_rate": 2.237585614215465e-05,
-      "loss": 1.8421,
-      "step": 757500
-    },
-    {
-      "epoch": 10.91,
-      "learning_rate": 2.2323929400401914e-05,
-      "loss": 1.8433,
-      "step": 758000
-    },
-    {
-      "epoch": 10.92,
-      "learning_rate": 2.227200265864918e-05,
-      "loss": 1.8404,
-      "step": 758500
-    },
-    {
-      "epoch": 10.93,
-      "learning_rate": 2.2220075916896444e-05,
-      "loss": 1.8417,
-      "step": 759000
-    },
-    {
-      "epoch": 10.93,
-      "learning_rate": 2.2168149175143706e-05,
-      "loss": 1.8366,
-      "step": 759500
-    },
-    {
-      "epoch": 10.94,
-      "learning_rate": 2.2116222433390974e-05,
-      "loss": 1.8378,
-      "step": 760000
-    },
-    {
-      "epoch": 10.95,
-      "learning_rate": 2.2064399545121744e-05,
-      "loss": 1.84,
-      "step": 760500
-    },
-    {
-      "epoch": 10.95,
-      "learning_rate": 2.201247280336901e-05,
-      "loss": 1.8389,
-      "step": 761000
-    },
-    {
-      "epoch": 10.96,
-      "learning_rate": 2.1960546061616273e-05,
-      "loss": 1.8398,
-      "step": 761500
-    },
-    {
-      "epoch": 10.97,
-      "learning_rate": 2.1908619319863538e-05,
-      "loss": 1.8433,
-      "step": 762000
-    },
-    {
-      "epoch": 10.98,
-      "learning_rate": 2.1856692578110803e-05,
-      "loss": 1.845,
-      "step": 762500
-    },
-    {
-      "epoch": 10.98,
-      "learning_rate": 2.1804869689841573e-05,
-      "loss": 1.8387,
-      "step": 763000
-    },
-    {
-      "epoch": 10.99,
-      "learning_rate": 2.1752942948088838e-05,
-      "loss": 1.8402,
-      "step": 763500
-    },
-    {
-      "epoch": 11.0,
-      "learning_rate": 2.1701016206336103e-05,
-      "loss": 1.8353,
-      "step": 764000
-    },
-    {
-      "epoch": 11.0,
-      "eval_accuracy": 0.6480513749434211,
-      "eval_loss": 1.702171802520752,
-      "eval_runtime": 647.7205,
-      "eval_samples_per_second": 832.052,
-      "eval_steps_per_second": 34.669,
-      "step": 764203
-    },
-    {
-      "epoch": 11.0,
-      "learning_rate": 2.1649089464583368e-05,
-      "loss": 1.837,
-      "step": 764500
-    },
-    {
-      "epoch": 11.01,
-      "learning_rate": 2.1597266576314137e-05,
-      "loss": 1.8367,
-      "step": 765000
-    },
-    {
-      "epoch": 11.02,
-      "learning_rate": 2.1545339834561402e-05,
-      "loss": 1.8372,
-      "step": 765500
-    },
-    {
-      "epoch": 11.03,
-      "learning_rate": 2.1493413092808667e-05,
-      "loss": 1.8376,
-      "step": 766000
-    },
-    {
-      "epoch": 11.03,
-      "learning_rate": 2.1441486351055932e-05,
-      "loss": 1.8358,
-      "step": 766500
-    },
-    {
-      "epoch": 11.04,
-      "learning_rate": 2.13896634627867e-05,
-      "loss": 1.8376,
-      "step": 767000
-    },
-    {
-      "epoch": 11.05,
-      "learning_rate": 2.1337736721033966e-05,
-      "loss": 1.8382,
-      "step": 767500
-    },
-    {
-      "epoch": 11.05,
-      "learning_rate": 2.128580997928123e-05,
-      "loss": 1.8359,
-      "step": 768000
-    },
-    {
-      "epoch": 11.06,
-      "learning_rate": 2.1233883237528496e-05,
-      "loss": 1.8354,
-      "step": 768500
-    },
-    {
-      "epoch": 11.07,
-      "learning_rate": 2.1181956495775758e-05,
-      "loss": 1.8305,
-      "step": 769000
-    },
-    {
-      "epoch": 11.08,
-      "learning_rate": 2.1130029754023026e-05,
-      "loss": 1.8408,
-      "step": 769500
-    },
-    {
-      "epoch": 11.08,
-      "learning_rate": 2.107810301227029e-05,
-      "loss": 1.8375,
-      "step": 770000
-    },
-    {
-      "epoch": 11.09,
-      "learning_rate": 2.1026176270517556e-05,
-      "loss": 1.8367,
-      "step": 770500
-    },
-    {
-      "epoch": 11.1,
-      "learning_rate": 2.0974353382248325e-05,
-      "loss": 1.8384,
-      "step": 771000
-    },
-    {
-      "epoch": 11.11,
-      "learning_rate": 2.092242664049559e-05,
-      "loss": 1.8399,
-      "step": 771500
-    },
-    {
-      "epoch": 11.11,
-      "learning_rate": 2.0870499898742855e-05,
-      "loss": 1.8376,
-      "step": 772000
-    },
-    {
-      "epoch": 11.12,
-      "learning_rate": 2.081857315699012e-05,
-      "loss": 1.8338,
-      "step": 772500
-    },
-    {
-      "epoch": 11.13,
-      "learning_rate": 2.0766646415237385e-05,
-      "loss": 1.839,
-      "step": 773000
-    },
-    {
-      "epoch": 11.13,
-      "learning_rate": 2.0714823526968155e-05,
-      "loss": 1.8359,
-      "step": 773500
-    },
-    {
-      "epoch": 11.14,
-      "learning_rate": 2.066289678521542e-05,
-      "loss": 1.8366,
-      "step": 774000
-    },
-    {
-      "epoch": 11.15,
-      "learning_rate": 2.0610970043462685e-05,
-      "loss": 1.8367,
-      "step": 774500
-    },
-    {
-      "epoch": 11.16,
-      "learning_rate": 2.055904330170995e-05,
-      "loss": 1.8327,
-      "step": 775000
-    },
-    {
-      "epoch": 11.16,
-      "learning_rate": 2.050711655995721e-05,
-      "loss": 1.8384,
-      "step": 775500
-    },
-    {
-      "epoch": 11.17,
-      "learning_rate": 2.0455293671687984e-05,
-      "loss": 1.839,
-      "step": 776000
-    },
-    {
-      "epoch": 11.18,
-      "learning_rate": 2.0403366929935245e-05,
-      "loss": 1.8345,
-      "step": 776500
-    },
-    {
-      "epoch": 11.18,
-      "learning_rate": 2.0351440188182514e-05,
-      "loss": 1.8379,
-      "step": 777000
-    },
-    {
-      "epoch": 11.19,
-      "learning_rate": 2.029951344642978e-05,
-      "loss": 1.834,
-      "step": 777500
-    },
-    {
-      "epoch": 11.2,
-      "learning_rate": 2.0247586704677044e-05,
-      "loss": 1.8389,
-      "step": 778000
-    },
-    {
-      "epoch": 11.21,
-      "learning_rate": 2.0195763816407813e-05,
-      "loss": 1.8309,
-      "step": 778500
-    },
-    {
-      "epoch": 11.21,
-      "learning_rate": 2.0143837074655078e-05,
-      "loss": 1.8367,
-      "step": 779000
-    },
-    {
-      "epoch": 11.22,
-      "learning_rate": 2.0091910332902343e-05,
-      "loss": 1.8367,
-      "step": 779500
-    },
-    {
-      "epoch": 11.23,
-      "learning_rate": 2.0039983591149608e-05,
-      "loss": 1.834,
-      "step": 780000
-    },
-    {
-      "epoch": 11.23,
-      "learning_rate": 1.9988056849396873e-05,
-      "loss": 1.837,
-      "step": 780500
-    },
-    {
-      "epoch": 11.24,
-      "learning_rate": 1.9936130107644134e-05,
-      "loss": 1.8344,
-      "step": 781000
-    },
-    {
-      "epoch": 11.25,
-      "learning_rate": 1.9884203365891403e-05,
-      "loss": 1.8388,
-      "step": 781500
-    },
-    {
-      "epoch": 11.26,
-      "learning_rate": 1.9832380477622172e-05,
-      "loss": 1.8317,
-      "step": 782000
-    },
-    {
-      "epoch": 11.26,
-      "learning_rate": 1.9780453735869437e-05,
-      "loss": 1.8334,
-      "step": 782500
-    },
-    {
-      "epoch": 11.27,
-      "learning_rate": 1.97285269941167e-05,
-      "loss": 1.8407,
-      "step": 783000
-    },
-    {
-      "epoch": 11.28,
-      "learning_rate": 1.9676600252363967e-05,
-      "loss": 1.8296,
-      "step": 783500
-    },
-    {
-      "epoch": 11.28,
-      "learning_rate": 1.962467351061123e-05,
-      "loss": 1.8319,
-      "step": 784000
-    },
-    {
-      "epoch": 11.29,
-      "learning_rate": 1.9572850622342e-05,
-      "loss": 1.8387,
-      "step": 784500
-    },
-    {
-      "epoch": 11.3,
-      "learning_rate": 1.9520923880589266e-05,
-      "loss": 1.8324,
-      "step": 785000
-    },
-    {
-      "epoch": 11.31,
-      "learning_rate": 1.946899713883653e-05,
-      "loss": 1.83,
-      "step": 785500
-    },
-    {
-      "epoch": 11.31,
-      "learning_rate": 1.9417070397083796e-05,
-      "loss": 1.8346,
-      "step": 786000
-    },
-    {
-      "epoch": 11.32,
-      "learning_rate": 1.936514365533106e-05,
-      "loss": 1.832,
-      "step": 786500
-    },
-    {
-      "epoch": 11.33,
-      "learning_rate": 1.9313216913578326e-05,
-      "loss": 1.8318,
-      "step": 787000
-    },
-    {
-      "epoch": 11.34,
-      "learning_rate": 1.9261290171825588e-05,
-      "loss": 1.8377,
-      "step": 787500
-    },
-    {
-      "epoch": 11.34,
-      "learning_rate": 1.920946728355636e-05,
-      "loss": 1.8335,
-      "step": 788000
-    },
-    {
-      "epoch": 11.35,
-      "learning_rate": 1.9157540541803622e-05,
-      "loss": 1.8374,
-      "step": 788500
-    },
-    {
-      "epoch": 11.36,
-      "learning_rate": 1.910561380005089e-05,
-      "loss": 1.8371,
-      "step": 789000
-    },
-    {
-      "epoch": 11.36,
-      "learning_rate": 1.9053687058298152e-05,
-      "loss": 1.8323,
-      "step": 789500
-    },
-    {
-      "epoch": 11.37,
-      "learning_rate": 1.900176031654542e-05,
-      "loss": 1.835,
-      "step": 790000
-    },
-    {
-      "epoch": 11.38,
-      "learning_rate": 1.8949937428276186e-05,
-      "loss": 1.8342,
-      "step": 790500
-    },
-    {
-      "epoch": 11.39,
-      "learning_rate": 1.8898010686523455e-05,
-      "loss": 1.8323,
-      "step": 791000
-    },
-    {
-      "epoch": 11.39,
-      "learning_rate": 1.8846083944770716e-05,
-      "loss": 1.8341,
-      "step": 791500
-    },
-    {
-      "epoch": 11.4,
-      "learning_rate": 1.8794157203017984e-05,
-      "loss": 1.8349,
-      "step": 792000
-    },
-    {
-      "epoch": 11.41,
-      "learning_rate": 1.874223046126525e-05,
-      "loss": 1.8314,
-      "step": 792500
-    },
-    {
-      "epoch": 11.41,
-      "learning_rate": 1.869040757299602e-05,
-      "loss": 1.8342,
-      "step": 793000
-    },
-    {
-      "epoch": 11.42,
-      "learning_rate": 1.8638480831243284e-05,
-      "loss": 1.8299,
-      "step": 793500
-    },
-    {
-      "epoch": 11.43,
-      "learning_rate": 1.858655408949055e-05,
-      "loss": 1.8336,
-      "step": 794000
-    },
-    {
-      "epoch": 11.44,
-      "learning_rate": 1.8534627347737814e-05,
-      "loss": 1.8342,
-      "step": 794500
-    },
-    {
-      "epoch": 11.44,
-      "learning_rate": 1.8482700605985075e-05,
-      "loss": 1.8341,
-      "step": 795000
-    },
-    {
-      "epoch": 11.45,
-      "learning_rate": 1.8430877717715848e-05,
-      "loss": 1.8299,
-      "step": 795500
-    },
-    {
-      "epoch": 11.46,
-      "learning_rate": 1.8378950975963113e-05,
-      "loss": 1.8331,
-      "step": 796000
-    },
-    {
-      "epoch": 11.46,
-      "learning_rate": 1.8327024234210378e-05,
-      "loss": 1.8351,
-      "step": 796500
-    },
-    {
-      "epoch": 11.47,
-      "learning_rate": 1.827509749245764e-05,
-      "loss": 1.8326,
-      "step": 797000
-    },
-    {
-      "epoch": 11.48,
-      "learning_rate": 1.8223170750704908e-05,
-      "loss": 1.8316,
-      "step": 797500
-    },
-    {
-      "epoch": 11.49,
-      "learning_rate": 1.8171347862435674e-05,
-      "loss": 1.8373,
-      "step": 798000
-    },
-    {
-      "epoch": 11.49,
-      "learning_rate": 1.8119421120682942e-05,
-      "loss": 1.8335,
-      "step": 798500
-    },
-    {
-      "epoch": 11.5,
-      "learning_rate": 1.8067494378930204e-05,
-      "loss": 1.8332,
-      "step": 799000
-    },
-    {
-      "epoch": 11.51,
-      "learning_rate": 1.8015567637177472e-05,
-      "loss": 1.8335,
-      "step": 799500
-    },
-    {
-      "epoch": 11.52,
-      "learning_rate": 1.7963640895424734e-05,
-      "loss": 1.834,
-      "step": 800000
-    },
-    {
-      "epoch": 11.52,
-      "learning_rate": 1.7911818007155507e-05,
-      "loss": 1.8336,
-      "step": 800500
-    },
-    {
-      "epoch": 11.53,
-      "learning_rate": 1.785989126540277e-05,
-      "loss": 1.8333,
-      "step": 801000
-    },
-    {
-      "epoch": 11.54,
-      "learning_rate": 1.7807964523650037e-05,
-      "loss": 1.8331,
-      "step": 801500
-    },
-    {
-      "epoch": 11.54,
-      "learning_rate": 1.77560377818973e-05,
-      "loss": 1.8332,
-      "step": 802000
-    },
-    {
-      "epoch": 11.55,
-      "learning_rate": 1.7704111040144563e-05,
-      "loss": 1.831,
-      "step": 802500
-    },
-    {
-      "epoch": 11.56,
-      "learning_rate": 1.765218429839183e-05,
-      "loss": 1.8317,
-      "step": 803000
-    },
-    {
-      "epoch": 11.57,
-      "learning_rate": 1.7600257556639093e-05,
-      "loss": 1.8344,
-      "step": 803500
-    },
-    {
-      "epoch": 11.57,
-      "learning_rate": 1.7548434668369866e-05,
-      "loss": 1.8334,
-      "step": 804000
-    },
-    {
-      "epoch": 11.58,
-      "learning_rate": 1.7496507926617127e-05,
-      "loss": 1.8333,
-      "step": 804500
-    },
-    {
-      "epoch": 11.59,
-      "learning_rate": 1.7444581184864396e-05,
-      "loss": 1.8317,
-      "step": 805000
-    },
-    {
-      "epoch": 11.59,
-      "learning_rate": 1.7392654443111657e-05,
-      "loss": 1.8368,
-      "step": 805500
-    },
-    {
-      "epoch": 11.6,
-      "learning_rate": 1.734083155484243e-05,
-      "loss": 1.8312,
-      "step": 806000
-    },
-    {
-      "epoch": 11.61,
-      "learning_rate": 1.728890481308969e-05,
-      "loss": 1.8325,
-      "step": 806500
-    },
-    {
-      "epoch": 11.62,
-      "learning_rate": 1.723697807133696e-05,
-      "loss": 1.8306,
-      "step": 807000
-    },
-    {
-      "epoch": 11.62,
-      "learning_rate": 1.718505132958422e-05,
-      "loss": 1.8335,
-      "step": 807500
-    },
-    {
-      "epoch": 11.63,
-      "learning_rate": 1.7133228441314994e-05,
-      "loss": 1.833,
-      "step": 808000
-    },
-    {
-      "epoch": 11.64,
-      "learning_rate": 1.7081301699562256e-05,
-      "loss": 1.8344,
-      "step": 808500
-    },
-    {
-      "epoch": 11.64,
-      "learning_rate": 1.7029374957809524e-05,
-      "loss": 1.8356,
-      "step": 809000
-    },
-    {
-      "epoch": 11.65,
-      "learning_rate": 1.697744821605679e-05,
-      "loss": 1.8329,
-      "step": 809500
-    },
-    {
-      "epoch": 11.66,
-      "learning_rate": 1.692552147430405e-05,
-      "loss": 1.8343,
-      "step": 810000
-    },
-    {
-      "epoch": 11.67,
-      "learning_rate": 1.6873698586034824e-05,
-      "loss": 1.8342,
-      "step": 810500
-    },
-    {
-      "epoch": 11.67,
-      "learning_rate": 1.682177184428209e-05,
-      "loss": 1.8328,
-      "step": 811000
-    },
-    {
-      "epoch": 11.68,
-      "learning_rate": 1.6769845102529354e-05,
-      "loss": 1.8328,
-      "step": 811500
-    },
-    {
-      "epoch": 11.69,
-      "learning_rate": 1.6717918360776615e-05,
-      "loss": 1.832,
-      "step": 812000
-    },
-    {
-      "epoch": 11.7,
-      "learning_rate": 1.6665991619023883e-05,
-      "loss": 1.8305,
-      "step": 812500
-    },
-    {
-      "epoch": 11.7,
-      "learning_rate": 1.6614064877271145e-05,
-      "loss": 1.8287,
-      "step": 813000
-    },
-    {
-      "epoch": 11.71,
-      "learning_rate": 1.6562241989001918e-05,
-      "loss": 1.8324,
-      "step": 813500
-    },
-    {
-      "epoch": 11.72,
-      "learning_rate": 1.651031524724918e-05,
-      "loss": 1.8325,
-      "step": 814000
-    },
-    {
-      "epoch": 11.72,
-      "learning_rate": 1.6458388505496448e-05,
-      "loss": 1.8304,
-      "step": 814500
-    },
-    {
-      "epoch": 11.73,
-      "learning_rate": 1.640646176374371e-05,
-      "loss": 1.8321,
-      "step": 815000
-    },
-    {
-      "epoch": 11.74,
-      "learning_rate": 1.6354638875474482e-05,
-      "loss": 1.8324,
-      "step": 815500
-    },
-    {
-      "epoch": 11.75,
-      "learning_rate": 1.6302712133721744e-05,
-      "loss": 1.8286,
-      "step": 816000
-    },
-    {
-      "epoch": 11.75,
-      "learning_rate": 1.6250785391969012e-05,
-      "loss": 1.8319,
-      "step": 816500
-    },
-    {
-      "epoch": 11.76,
-      "learning_rate": 1.6198858650216277e-05,
-      "loss": 1.8282,
-      "step": 817000
-    },
-    {
-      "epoch": 11.77,
-      "learning_rate": 1.6147035761947047e-05,
-      "loss": 1.8265,
-      "step": 817500
-    },
-    {
-      "epoch": 11.77,
-      "learning_rate": 1.609510902019431e-05,
-      "loss": 1.8288,
-      "step": 818000
-    },
-    {
-      "epoch": 11.78,
-      "learning_rate": 1.6043182278441576e-05,
-      "loss": 1.8334,
-      "step": 818500
-    },
-    {
-      "epoch": 11.79,
-      "learning_rate": 1.599125553668884e-05,
-      "loss": 1.8322,
-      "step": 819000
-    },
-    {
-      "epoch": 11.8,
-      "learning_rate": 1.5939328794936103e-05,
-      "loss": 1.8318,
-      "step": 819500
-    },
-    {
-      "epoch": 11.8,
-      "learning_rate": 1.5887505906666876e-05,
-      "loss": 1.8281,
-      "step": 820000
-    },
-    {
-      "epoch": 11.81,
-      "learning_rate": 1.583557916491414e-05,
-      "loss": 1.8279,
-      "step": 820500
-    },
-    {
-      "epoch": 11.82,
-      "learning_rate": 1.5783652423161406e-05,
-      "loss": 1.8319,
-      "step": 821000
-    },
-    {
-      "epoch": 11.82,
-      "learning_rate": 1.5731725681408667e-05,
-      "loss": 1.8281,
-      "step": 821500
-    },
-    {
-      "epoch": 11.83,
-      "learning_rate": 1.5679798939655935e-05,
-      "loss": 1.83,
-      "step": 822000
-    },
-    {
-      "epoch": 11.84,
-      "learning_rate": 1.5627976051386705e-05,
-      "loss": 1.831,
-      "step": 822500
-    },
-    {
-      "epoch": 11.85,
-      "learning_rate": 1.557604930963397e-05,
-      "loss": 1.8308,
-      "step": 823000
-    },
-    {
-      "epoch": 11.85,
-      "learning_rate": 1.552412256788123e-05,
-      "loss": 1.8302,
-      "step": 823500
-    },
-    {
-      "epoch": 11.86,
-      "learning_rate": 1.54721958261285e-05,
-      "loss": 1.8306,
-      "step": 824000
-    },
-    {
-      "epoch": 11.87,
-      "learning_rate": 1.5420372937859266e-05,
-      "loss": 1.828,
-      "step": 824500
-    },
-    {
-      "epoch": 11.88,
-      "learning_rate": 1.5368446196106534e-05,
-      "loss": 1.8306,
-      "step": 825000
-    },
-    {
-      "epoch": 11.88,
-      "learning_rate": 1.53165194543538e-05,
-      "loss": 1.8288,
-      "step": 825500
-    },
-    {
-      "epoch": 11.89,
-      "learning_rate": 1.5264592712601064e-05,
-      "loss": 1.8304,
-      "step": 826000
-    },
-    {
-      "epoch": 11.9,
-      "learning_rate": 1.5212769824331832e-05,
-      "loss": 1.8285,
-      "step": 826500
-    },
-    {
-      "epoch": 11.9,
-      "learning_rate": 1.5160843082579099e-05,
-      "loss": 1.8287,
-      "step": 827000
-    },
-    {
-      "epoch": 11.91,
-      "learning_rate": 1.5108916340826362e-05,
-      "loss": 1.8291,
-      "step": 827500
-    },
-    {
-      "epoch": 11.92,
-      "learning_rate": 1.5056989599073628e-05,
-      "loss": 1.8314,
-      "step": 828000
-    },
-    {
-      "epoch": 11.93,
-      "learning_rate": 1.5005062857320892e-05,
-      "loss": 1.8258,
-      "step": 828500
-    },
-    {
-      "epoch": 11.93,
-      "learning_rate": 1.4953239969051663e-05,
-      "loss": 1.8287,
-      "step": 829000
-    },
-    {
-      "epoch": 11.94,
-      "learning_rate": 1.4901313227298926e-05,
-      "loss": 1.8294,
-      "step": 829500
-    },
-    {
-      "epoch": 11.95,
-      "learning_rate": 1.4849386485546193e-05,
-      "loss": 1.8328,
-      "step": 830000
-    },
-    {
-      "epoch": 11.95,
-      "learning_rate": 1.4797459743793458e-05,
-      "loss": 1.8287,
-      "step": 830500
-    },
-    {
-      "epoch": 11.96,
-      "learning_rate": 1.4745636855524227e-05,
-      "loss": 1.8279,
-      "step": 831000
-    },
-    {
-      "epoch": 11.97,
-      "learning_rate": 1.4693710113771492e-05,
-      "loss": 1.8286,
-      "step": 831500
-    },
-    {
-      "epoch": 11.98,
-      "learning_rate": 1.4641783372018755e-05,
-      "loss": 1.8254,
-      "step": 832000
-    },
-    {
-      "epoch": 11.98,
-      "learning_rate": 1.4589856630266022e-05,
-      "loss": 1.8274,
-      "step": 832500
-    },
-    {
-      "epoch": 11.99,
-      "learning_rate": 1.4538033741996793e-05,
-      "loss": 1.8279,
-      "step": 833000
-    },
-    {
-      "epoch": 12.0,
-      "learning_rate": 1.4486107000244057e-05,
-      "loss": 1.8245,
-      "step": 833500
-    },
-    {
-      "epoch": 12.0,
-      "eval_accuracy": 0.6503763935317559,
-      "eval_loss": 1.6876965761184692,
-      "eval_runtime": 647.389,
-      "eval_samples_per_second": 832.478,
-      "eval_steps_per_second": 34.687,
-      "step": 833676
-    },
-    {
-      "epoch": 12.0,
-      "learning_rate": 1.443418025849132e-05,
-      "loss": 1.8307,
-      "step": 834000
-    },
-    {
-      "epoch": 12.01,
-      "learning_rate": 1.4382253516738586e-05,
-      "loss": 1.8256,
-      "step": 834500
-    },
-    {
-      "epoch": 12.02,
-      "learning_rate": 1.4330430628469358e-05,
-      "loss": 1.829,
-      "step": 835000
-    },
-    {
-      "epoch": 12.03,
-      "learning_rate": 1.4278503886716621e-05,
-      "loss": 1.8268,
-      "step": 835500
-    },
-    {
-      "epoch": 12.03,
-      "learning_rate": 1.4226577144963884e-05,
-      "loss": 1.8277,
-      "step": 836000
-    },
-    {
-      "epoch": 12.04,
-      "learning_rate": 1.417465040321115e-05,
-      "loss": 1.8258,
-      "step": 836500
-    },
-    {
-      "epoch": 12.05,
-      "learning_rate": 1.4122723661458414e-05,
-      "loss": 1.8224,
-      "step": 837000
-    },
-    {
-      "epoch": 12.06,
-      "learning_rate": 1.4070900773189185e-05,
-      "loss": 1.827,
-      "step": 837500
-    },
-    {
-      "epoch": 12.06,
-      "learning_rate": 1.401897403143645e-05,
-      "loss": 1.8239,
-      "step": 838000
-    },
-    {
-      "epoch": 12.07,
-      "learning_rate": 1.3967047289683715e-05,
-      "loss": 1.8255,
-      "step": 838500
-    },
-    {
-      "epoch": 12.08,
-      "learning_rate": 1.391512054793098e-05,
-      "loss": 1.8282,
-      "step": 839000
-    },
-    {
-      "epoch": 12.08,
-      "learning_rate": 1.3863193806178247e-05,
-      "loss": 1.8262,
-      "step": 839500
-    },
-    {
-      "epoch": 12.09,
-      "learning_rate": 1.3811370917909014e-05,
-      "loss": 1.8251,
-      "step": 840000
-    },
-    {
-      "epoch": 12.1,
-      "learning_rate": 1.3759444176156281e-05,
-      "loss": 1.8257,
-      "step": 840500
-    },
-    {
-      "epoch": 12.11,
-      "learning_rate": 1.3707517434403544e-05,
-      "loss": 1.825,
-      "step": 841000
-    },
-    {
-      "epoch": 12.11,
-      "learning_rate": 1.3655590692650807e-05,
-      "loss": 1.8269,
-      "step": 841500
-    },
-    {
-      "epoch": 12.12,
-      "learning_rate": 1.3603767804381579e-05,
-      "loss": 1.8266,
-      "step": 842000
-    },
-    {
-      "epoch": 12.13,
-      "learning_rate": 1.3551841062628845e-05,
-      "loss": 1.8275,
-      "step": 842500
-    },
-    {
-      "epoch": 12.13,
-      "learning_rate": 1.3499914320876109e-05,
-      "loss": 1.8266,
-      "step": 843000
-    },
-    {
-      "epoch": 12.14,
-      "learning_rate": 1.3447987579123372e-05,
-      "loss": 1.8256,
-      "step": 843500
-    },
-    {
-      "epoch": 12.15,
-      "learning_rate": 1.3396164690854143e-05,
-      "loss": 1.8266,
-      "step": 844000
-    },
-    {
-      "epoch": 12.16,
-      "learning_rate": 1.334423794910141e-05,
-      "loss": 1.8231,
-      "step": 844500
-    },
-    {
-      "epoch": 12.16,
-      "learning_rate": 1.3292311207348673e-05,
-      "loss": 1.8236,
-      "step": 845000
-    },
-    {
-      "epoch": 12.17,
-      "learning_rate": 1.3240384465595936e-05,
-      "loss": 1.8239,
-      "step": 845500
-    },
-    {
-      "epoch": 12.18,
-      "learning_rate": 1.3188561577326707e-05,
-      "loss": 1.8254,
-      "step": 846000
-    },
-    {
-      "epoch": 12.18,
-      "learning_rate": 1.3136634835573972e-05,
-      "loss": 1.8244,
-      "step": 846500
-    },
-    {
-      "epoch": 12.19,
-      "learning_rate": 1.3084708093821237e-05,
-      "loss": 1.8238,
-      "step": 847000
-    },
-    {
-      "epoch": 12.2,
-      "learning_rate": 1.3032781352068502e-05,
-      "loss": 1.823,
-      "step": 847500
-    },
-    {
-      "epoch": 12.21,
-      "learning_rate": 1.2980854610315769e-05,
-      "loss": 1.8253,
-      "step": 848000
-    },
-    {
-      "epoch": 12.21,
-      "learning_rate": 1.2929031722046537e-05,
-      "loss": 1.8226,
-      "step": 848500
-    },
-    {
-      "epoch": 12.22,
-      "learning_rate": 1.2877104980293803e-05,
-      "loss": 1.8251,
-      "step": 849000
-    },
-    {
-      "epoch": 12.23,
-      "learning_rate": 1.2825178238541067e-05,
-      "loss": 1.8232,
-      "step": 849500
-    },
-    {
-      "epoch": 12.23,
-      "learning_rate": 1.2773251496788333e-05,
-      "loss": 1.8279,
-      "step": 850000
-    },
-    {
-      "epoch": 12.24,
-      "learning_rate": 1.2721324755035596e-05,
-      "loss": 1.8273,
-      "step": 850500
-    },
-    {
-      "epoch": 12.25,
-      "learning_rate": 1.2669501866766368e-05,
-      "loss": 1.8245,
-      "step": 851000
-    },
-    {
-      "epoch": 12.26,
-      "learning_rate": 1.2617575125013631e-05,
-      "loss": 1.8226,
-      "step": 851500
-    },
-    {
-      "epoch": 12.26,
-      "learning_rate": 1.2565648383260897e-05,
-      "loss": 1.8262,
-      "step": 852000
-    },
-    {
-      "epoch": 12.27,
-      "learning_rate": 1.251372164150816e-05,
-      "loss": 1.8248,
-      "step": 852500
-    },
-    {
-      "epoch": 12.28,
-      "learning_rate": 1.246189875323893e-05,
-      "loss": 1.8203,
-      "step": 853000
-    },
-    {
-      "epoch": 12.29,
-      "learning_rate": 1.2409972011486195e-05,
-      "loss": 1.8257,
-      "step": 853500
-    },
-    {
-      "epoch": 12.29,
-      "learning_rate": 1.235804526973346e-05,
-      "loss": 1.8219,
-      "step": 854000
-    },
-    {
-      "epoch": 12.3,
-      "learning_rate": 1.2306118527980725e-05,
-      "loss": 1.8203,
-      "step": 854500
-    },
-    {
-      "epoch": 12.31,
-      "learning_rate": 1.2254295639711495e-05,
-      "loss": 1.8241,
-      "step": 855000
-    },
-    {
-      "epoch": 12.31,
-      "learning_rate": 1.220236889795876e-05,
-      "loss": 1.8251,
-      "step": 855500
-    },
-    {
-      "epoch": 12.32,
-      "learning_rate": 1.2150442156206024e-05,
-      "loss": 1.8244,
-      "step": 856000
-    },
-    {
-      "epoch": 12.33,
-      "learning_rate": 1.2098515414453291e-05,
-      "loss": 1.8227,
-      "step": 856500
-    },
-    {
-      "epoch": 12.34,
-      "learning_rate": 1.204669252618406e-05,
-      "loss": 1.8229,
-      "step": 857000
-    },
-    {
-      "epoch": 12.34,
-      "learning_rate": 1.1994765784431326e-05,
-      "loss": 1.8256,
-      "step": 857500
-    },
-    {
-      "epoch": 12.35,
-      "learning_rate": 1.194283904267859e-05,
-      "loss": 1.8224,
-      "step": 858000
-    },
-    {
-      "epoch": 12.36,
-      "learning_rate": 1.1890912300925855e-05,
-      "loss": 1.8221,
-      "step": 858500
-    },
-    {
-      "epoch": 12.36,
-      "learning_rate": 1.1838985559173119e-05,
-      "loss": 1.826,
-      "step": 859000
-    },
-    {
-      "epoch": 12.37,
-      "learning_rate": 1.178716267090389e-05,
-      "loss": 1.8251,
-      "step": 859500
-    },
-    {
-      "epoch": 12.38,
-      "learning_rate": 1.1735235929151155e-05,
-      "loss": 1.8236,
-      "step": 860000
-    },
-    {
-      "epoch": 12.39,
-      "learning_rate": 1.1683309187398418e-05,
-      "loss": 1.82,
-      "step": 860500
-    },
-    {
-      "epoch": 12.39,
-      "learning_rate": 1.1631382445645683e-05,
-      "loss": 1.8277,
-      "step": 861000
-    },
-    {
-      "epoch": 12.4,
-      "learning_rate": 1.1579455703892948e-05,
-      "loss": 1.8275,
-      "step": 861500
-    },
-    {
-      "epoch": 12.41,
-      "learning_rate": 1.152763281562372e-05,
-      "loss": 1.8221,
-      "step": 862000
-    },
-    {
-      "epoch": 12.41,
-      "learning_rate": 1.1475706073870982e-05,
-      "loss": 1.8203,
-      "step": 862500
-    },
-    {
-      "epoch": 12.42,
-      "learning_rate": 1.1423779332118247e-05,
-      "loss": 1.8198,
-      "step": 863000
-    },
-    {
-      "epoch": 12.43,
-      "learning_rate": 1.1371852590365512e-05,
-      "loss": 1.8251,
-      "step": 863500
-    },
-    {
-      "epoch": 12.44,
-      "learning_rate": 1.1319925848612779e-05,
-      "loss": 1.8221,
-      "step": 864000
-    },
-    {
-      "epoch": 12.44,
-      "learning_rate": 1.1268102960343548e-05,
-      "loss": 1.8245,
-      "step": 864500
-    },
-    {
-      "epoch": 12.45,
-      "learning_rate": 1.1216176218590813e-05,
-      "loss": 1.8263,
-      "step": 865000
-    },
-    {
-      "epoch": 12.46,
-      "learning_rate": 1.1164249476838078e-05,
-      "loss": 1.8232,
-      "step": 865500
-    },
-    {
-      "epoch": 12.47,
-      "learning_rate": 1.1112322735085343e-05,
-      "loss": 1.821,
-      "step": 866000
-    },
-    {
-      "epoch": 12.47,
-      "learning_rate": 1.1060499846816113e-05,
-      "loss": 1.8223,
-      "step": 866500
-    },
-    {
-      "epoch": 12.48,
-      "learning_rate": 1.1008573105063378e-05,
-      "loss": 1.8272,
-      "step": 867000
-    },
-    {
-      "epoch": 12.49,
-      "learning_rate": 1.0956646363310643e-05,
-      "loss": 1.8216,
-      "step": 867500
-    },
-    {
-      "epoch": 12.49,
-      "learning_rate": 1.0904719621557907e-05,
-      "loss": 1.823,
-      "step": 868000
-    },
-    {
-      "epoch": 12.5,
-      "learning_rate": 1.085279287980517e-05,
-      "loss": 1.8194,
-      "step": 868500
-    },
-    {
-      "epoch": 12.51,
-      "learning_rate": 1.0800969991535942e-05,
-      "loss": 1.8213,
-      "step": 869000
-    },
-    {
-      "epoch": 12.52,
-      "learning_rate": 1.0749043249783207e-05,
-      "loss": 1.8196,
-      "step": 869500
-    },
-    {
-      "epoch": 12.52,
-      "learning_rate": 1.069711650803047e-05,
-      "loss": 1.8231,
-      "step": 870000
-    },
-    {
-      "epoch": 12.53,
-      "learning_rate": 1.0645189766277735e-05,
-      "loss": 1.8158,
-      "step": 870500
-    },
-    {
-      "epoch": 12.54,
-      "learning_rate": 1.0593366878008506e-05,
-      "loss": 1.8195,
-      "step": 871000
-    },
-    {
-      "epoch": 12.54,
-      "learning_rate": 1.054144013625577e-05,
-      "loss": 1.8227,
-      "step": 871500
-    },
-    {
-      "epoch": 12.55,
-      "learning_rate": 1.0489513394503034e-05,
-      "loss": 1.8157,
-      "step": 872000
-    },
-    {
-      "epoch": 12.56,
-      "learning_rate": 1.0437586652750301e-05,
-      "loss": 1.8217,
-      "step": 872500
-    },
-    {
-      "epoch": 12.57,
-      "learning_rate": 1.0385659910997566e-05,
-      "loss": 1.8213,
-      "step": 873000
-    },
-    {
-      "epoch": 12.57,
-      "learning_rate": 1.0333837022728336e-05,
-      "loss": 1.8229,
-      "step": 873500
-    },
-    {
-      "epoch": 12.58,
-      "learning_rate": 1.02819102809756e-05,
-      "loss": 1.8199,
-      "step": 874000
-    },
-    {
-      "epoch": 12.59,
-      "learning_rate": 1.0229983539222865e-05,
-      "loss": 1.8235,
-      "step": 874500
-    },
-    {
-      "epoch": 12.59,
-      "learning_rate": 1.017805679747013e-05,
-      "loss": 1.8226,
-      "step": 875000
-    },
-    {
-      "epoch": 12.6,
-      "learning_rate": 1.0126130055717395e-05,
-      "loss": 1.8207,
-      "step": 875500
-    },
-    {
-      "epoch": 12.61,
-      "learning_rate": 1.0074307167448165e-05,
-      "loss": 1.8171,
-      "step": 876000
-    },
-    {
-      "epoch": 12.62,
-      "learning_rate": 1.002238042569543e-05,
-      "loss": 1.8231,
-      "step": 876500
-    },
-    {
-      "epoch": 12.62,
-      "learning_rate": 9.970453683942695e-06,
-      "loss": 1.8192,
-      "step": 877000
-    },
-    {
-      "epoch": 12.63,
-      "learning_rate": 9.918526942189958e-06,
-      "loss": 1.8194,
-      "step": 877500
-    },
-    {
-      "epoch": 12.64,
-      "learning_rate": 9.866600200437223e-06,
-      "loss": 1.8203,
-      "step": 878000
-    },
-    {
-      "epoch": 12.65,
-      "learning_rate": 9.814673458684488e-06,
-      "loss": 1.8184,
-      "step": 878500
-    },
-    {
-      "epoch": 12.65,
-      "learning_rate": 9.762850570415259e-06,
-      "loss": 1.8189,
-      "step": 879000
-    },
-    {
-      "epoch": 12.66,
-      "learning_rate": 9.710923828662522e-06,
-      "loss": 1.8217,
-      "step": 879500
-    },
-    {
-      "epoch": 12.67,
-      "learning_rate": 9.658997086909787e-06,
-      "loss": 1.8219,
-      "step": 880000
-    },
-    {
-      "epoch": 12.67,
-      "learning_rate": 9.607070345157054e-06,
-      "loss": 1.8226,
-      "step": 880500
-    },
-    {
-      "epoch": 12.68,
-      "learning_rate": 9.555247456887823e-06,
-      "loss": 1.8191,
-      "step": 881000
-    },
-    {
-      "epoch": 12.69,
-      "learning_rate": 9.503320715135088e-06,
-      "loss": 1.8191,
-      "step": 881500
-    },
-    {
-      "epoch": 12.7,
-      "learning_rate": 9.451393973382353e-06,
-      "loss": 1.821,
-      "step": 882000
-    },
-    {
-      "epoch": 12.7,
-      "learning_rate": 9.399467231629618e-06,
-      "loss": 1.8177,
-      "step": 882500
-    },
-    {
-      "epoch": 12.71,
-      "learning_rate": 9.347644343360388e-06,
-      "loss": 1.8254,
-      "step": 883000
-    },
-    {
-      "epoch": 12.72,
-      "learning_rate": 9.295717601607653e-06,
-      "loss": 1.8205,
-      "step": 883500
-    },
-    {
-      "epoch": 12.72,
-      "learning_rate": 9.243790859854917e-06,
-      "loss": 1.8196,
-      "step": 884000
-    },
-    {
-      "epoch": 12.73,
-      "learning_rate": 9.191864118102182e-06,
-      "loss": 1.8206,
-      "step": 884500
-    },
-    {
-      "epoch": 12.74,
-      "learning_rate": 9.139937376349447e-06,
-      "loss": 1.8201,
-      "step": 885000
-    },
-    {
-      "epoch": 12.75,
-      "learning_rate": 9.088114488080217e-06,
-      "loss": 1.8182,
-      "step": 885500
-    },
-    {
-      "epoch": 12.75,
-      "learning_rate": 9.036187746327482e-06,
-      "loss": 1.8198,
-      "step": 886000
-    },
-    {
-      "epoch": 12.76,
-      "learning_rate": 8.984261004574747e-06,
-      "loss": 1.8238,
-      "step": 886500
-    },
-    {
-      "epoch": 12.77,
-      "learning_rate": 8.93233426282201e-06,
-      "loss": 1.8186,
-      "step": 887000
-    },
-    {
-      "epoch": 12.77,
-      "learning_rate": 8.880407521069275e-06,
-      "loss": 1.8159,
-      "step": 887500
-    },
-    {
-      "epoch": 12.78,
-      "learning_rate": 8.828584632800046e-06,
-      "loss": 1.818,
-      "step": 888000
-    },
-    {
-      "epoch": 12.79,
-      "learning_rate": 8.776657891047311e-06,
-      "loss": 1.8221,
-      "step": 888500
-    },
-    {
-      "epoch": 12.8,
-      "learning_rate": 8.724731149294576e-06,
-      "loss": 1.8165,
-      "step": 889000
-    },
-    {
-      "epoch": 12.8,
-      "learning_rate": 8.67280440754184e-06,
-      "loss": 1.8205,
-      "step": 889500
-    },
-    {
-      "epoch": 12.81,
-      "learning_rate": 8.62098151927261e-06,
-      "loss": 1.8192,
-      "step": 890000
-    },
-    {
-      "epoch": 12.82,
-      "learning_rate": 8.569054777519875e-06,
-      "loss": 1.8155,
-      "step": 890500
-    },
-    {
-      "epoch": 12.83,
-      "learning_rate": 8.51712803576714e-06,
-      "loss": 1.8241,
-      "step": 891000
-    },
-    {
-      "epoch": 12.83,
-      "learning_rate": 8.465201294014405e-06,
-      "loss": 1.8203,
-      "step": 891500
-    },
-    {
-      "epoch": 12.84,
-      "learning_rate": 8.413378405745175e-06,
-      "loss": 1.8153,
-      "step": 892000
-    },
-    {
-      "epoch": 12.85,
-      "learning_rate": 8.36145166399244e-06,
-      "loss": 1.8188,
-      "step": 892500
-    },
-    {
-      "epoch": 12.85,
-      "learning_rate": 8.309524922239705e-06,
-      "loss": 1.8233,
-      "step": 893000
-    },
-    {
-      "epoch": 12.86,
-      "learning_rate": 8.25759818048697e-06,
-      "loss": 1.82,
-      "step": 893500
-    },
-    {
-      "epoch": 12.87,
-      "learning_rate": 8.205775292217739e-06,
-      "loss": 1.8259,
-      "step": 894000
-    },
-    {
-      "epoch": 12.88,
-      "learning_rate": 8.153848550465004e-06,
-      "loss": 1.8141,
-      "step": 894500
-    },
-    {
-      "epoch": 12.88,
-      "learning_rate": 8.101921808712269e-06,
-      "loss": 1.8185,
-      "step": 895000
-    },
-    {
-      "epoch": 12.89,
-      "learning_rate": 8.049995066959534e-06,
-      "loss": 1.8158,
-      "step": 895500
-    },
-    {
-      "epoch": 12.9,
-      "learning_rate": 7.998172178690303e-06,
-      "loss": 1.8182,
-      "step": 896000
-    },
-    {
-      "epoch": 12.9,
-      "learning_rate": 7.946245436937568e-06,
-      "loss": 1.8226,
-      "step": 896500
-    },
-    {
-      "epoch": 12.91,
-      "learning_rate": 7.894318695184833e-06,
-      "loss": 1.82,
-      "step": 897000
-    },
-    {
-      "epoch": 12.92,
-      "learning_rate": 7.842391953432098e-06,
-      "loss": 1.8245,
-      "step": 897500
-    },
-    {
-      "epoch": 12.93,
-      "learning_rate": 7.790465211679363e-06,
-      "loss": 1.8168,
-      "step": 898000
-    },
-    {
-      "epoch": 12.93,
-      "learning_rate": 7.738538469926628e-06,
-      "loss": 1.8208,
-      "step": 898500
-    },
-    {
-      "epoch": 12.94,
-      "learning_rate": 7.686611728173893e-06,
-      "loss": 1.8174,
-      "step": 899000
-    },
-    {
-      "epoch": 12.95,
-      "learning_rate": 7.634788839904664e-06,
-      "loss": 1.8105,
-      "step": 899500
-    },
-    {
-      "epoch": 12.95,
-      "learning_rate": 7.5828620981519274e-06,
-      "loss": 1.8165,
-      "step": 900000
-    },
-    {
-      "epoch": 12.96,
-      "learning_rate": 7.530935356399192e-06,
-      "loss": 1.8201,
-      "step": 900500
-    },
-    {
-      "epoch": 12.97,
-      "learning_rate": 7.479008614646457e-06,
-      "loss": 1.8207,
-      "step": 901000
-    },
-    {
-      "epoch": 12.98,
-      "learning_rate": 7.427185726377227e-06,
-      "loss": 1.8191,
-      "step": 901500
-    },
-    {
-      "epoch": 12.98,
-      "learning_rate": 7.375258984624492e-06,
-      "loss": 1.8161,
-      "step": 902000
-    },
-    {
-      "epoch": 12.99,
-      "learning_rate": 7.323332242871757e-06,
-      "loss": 1.8179,
-      "step": 902500
-    },
-    {
-      "epoch": 13.0,
-      "learning_rate": 7.271405501119022e-06,
-      "loss": 1.8191,
-      "step": 903000
-    },
-    {
-      "epoch": 13.0,
-      "eval_accuracy": 0.651487792294714,
-      "eval_loss": 1.6829075813293457,
-      "eval_runtime": 646.4178,
-      "eval_samples_per_second": 833.729,
-      "eval_steps_per_second": 34.739,
-      "step": 903149
-    },
-    {
-      "epoch": 13.01,
-      "learning_rate": 7.2194787593662865e-06,
-      "loss": 1.8161,
-      "step": 903500
-    },
-    {
-      "epoch": 13.01,
-      "learning_rate": 7.167655871097056e-06,
-      "loss": 1.8188,
-      "step": 904000
-    },
-    {
-      "epoch": 13.02,
-      "learning_rate": 7.115729129344322e-06,
-      "loss": 1.8164,
-      "step": 904500
-    },
-    {
-      "epoch": 13.03,
-      "learning_rate": 7.063802387591587e-06,
-      "loss": 1.8137,
-      "step": 905000
-    },
-    {
-      "epoch": 13.03,
-      "learning_rate": 7.011875645838852e-06,
-      "loss": 1.8147,
-      "step": 905500
-    },
-    {
-      "epoch": 13.04,
-      "learning_rate": 6.960052757569621e-06,
-      "loss": 1.8158,
-      "step": 906000
-    },
-    {
-      "epoch": 13.05,
-      "learning_rate": 6.908126015816886e-06,
-      "loss": 1.8141,
-      "step": 906500
-    },
-    {
-      "epoch": 13.06,
-      "learning_rate": 6.856199274064151e-06,
-      "loss": 1.8159,
-      "step": 907000
-    },
-    {
-      "epoch": 13.06,
-      "learning_rate": 6.804272532311415e-06,
-      "loss": 1.8189,
-      "step": 907500
-    },
-    {
-      "epoch": 13.07,
-      "learning_rate": 6.752449644042186e-06,
-      "loss": 1.818,
-      "step": 908000
-    },
-    {
-      "epoch": 13.08,
-      "learning_rate": 6.7005229022894505e-06,
-      "loss": 1.8195,
-      "step": 908500
-    },
-    {
-      "epoch": 13.08,
-      "learning_rate": 6.648596160536715e-06,
-      "loss": 1.8131,
-      "step": 909000
-    },
-    {
-      "epoch": 13.09,
-      "learning_rate": 6.5966694187839795e-06,
-      "loss": 1.8176,
-      "step": 909500
-    },
-    {
-      "epoch": 13.1,
-      "learning_rate": 6.544742677031244e-06,
-      "loss": 1.8162,
-      "step": 910000
-    },
-    {
-      "epoch": 13.11,
-      "learning_rate": 6.492919788762016e-06,
-      "loss": 1.8171,
-      "step": 910500
-    },
-    {
-      "epoch": 13.11,
-      "learning_rate": 6.440993047009279e-06,
-      "loss": 1.8127,
-      "step": 911000
-    },
-    {
-      "epoch": 13.12,
-      "learning_rate": 6.389066305256544e-06,
-      "loss": 1.8153,
-      "step": 911500
-    },
-    {
-      "epoch": 13.13,
-      "learning_rate": 6.337139563503809e-06,
-      "loss": 1.816,
-      "step": 912000
-    },
-    {
-      "epoch": 13.13,
-      "learning_rate": 6.2852128217510745e-06,
-      "loss": 1.8159,
-      "step": 912500
-    },
-    {
-      "epoch": 13.14,
-      "learning_rate": 6.233389933481844e-06,
-      "loss": 1.8191,
-      "step": 913000
-    },
-    {
-      "epoch": 13.15,
-      "learning_rate": 6.181463191729109e-06,
-      "loss": 1.8146,
-      "step": 913500
-    },
-    {
-      "epoch": 13.16,
-      "learning_rate": 6.129536449976374e-06,
-      "loss": 1.8127,
-      "step": 914000
-    },
-    {
-      "epoch": 13.16,
-      "learning_rate": 6.077609708223638e-06,
-      "loss": 1.8159,
-      "step": 914500
-    },
-    {
-      "epoch": 13.17,
-      "learning_rate": 6.0257868199544085e-06,
-      "loss": 1.8137,
-      "step": 915000
-    },
-    {
-      "epoch": 13.18,
-      "learning_rate": 5.973860078201673e-06,
-      "loss": 1.8139,
-      "step": 915500
-    },
-    {
-      "epoch": 13.18,
-      "learning_rate": 5.9219333364489374e-06,
-      "loss": 1.8123,
-      "step": 916000
-    },
-    {
-      "epoch": 13.19,
-      "learning_rate": 5.870006594696203e-06,
-      "loss": 1.8142,
-      "step": 916500
-    },
-    {
-      "epoch": 13.2,
-      "learning_rate": 5.818079852943468e-06,
-      "loss": 1.8136,
-      "step": 917000
-    },
-    {
-      "epoch": 13.21,
-      "learning_rate": 5.7662569646742386e-06,
-      "loss": 1.8203,
-      "step": 917500
-    },
-    {
-      "epoch": 13.21,
-      "learning_rate": 5.714330222921503e-06,
-      "loss": 1.8131,
-      "step": 918000
-    },
-    {
-      "epoch": 13.22,
-      "learning_rate": 5.6624034811687675e-06,
-      "loss": 1.814,
-      "step": 918500
-    },
-    {
-      "epoch": 13.23,
-      "learning_rate": 5.6104767394160316e-06,
-      "loss": 1.8151,
-      "step": 919000
-    },
-    {
-      "epoch": 13.24,
-      "learning_rate": 5.5585499976632965e-06,
-      "loss": 1.8176,
-      "step": 919500
-    },
-    {
-      "epoch": 13.24,
-      "learning_rate": 5.506727109394067e-06,
-      "loss": 1.8139,
-      "step": 920000
-    },
-    {
-      "epoch": 13.25,
-      "learning_rate": 5.454800367641332e-06,
-      "loss": 1.8159,
-      "step": 920500
-    },
-    {
-      "epoch": 13.26,
-      "learning_rate": 5.402873625888597e-06,
-      "loss": 1.8192,
-      "step": 921000
-    },
-    {
-      "epoch": 13.26,
-      "learning_rate": 5.350946884135862e-06,
-      "loss": 1.8116,
-      "step": 921500
-    },
-    {
-      "epoch": 13.27,
-      "learning_rate": 5.299123995866632e-06,
-      "loss": 1.8147,
-      "step": 922000
-    },
-    {
-      "epoch": 13.28,
-      "learning_rate": 5.247197254113896e-06,
-      "loss": 1.812,
-      "step": 922500
-    },
-    {
-      "epoch": 13.29,
-      "learning_rate": 5.195270512361161e-06,
-      "loss": 1.8176,
-      "step": 923000
-    },
-    {
-      "epoch": 13.29,
-      "learning_rate": 5.143343770608426e-06,
-      "loss": 1.814,
-      "step": 923500
-    },
-    {
-      "epoch": 13.3,
-      "learning_rate": 5.091520882339196e-06,
-      "loss": 1.8153,
-      "step": 924000
-    },
-    {
-      "epoch": 13.31,
-      "learning_rate": 5.0395941405864605e-06,
-      "loss": 1.8171,
-      "step": 924500
-    },
-    {
-      "epoch": 13.31,
-      "learning_rate": 4.9876673988337254e-06,
-      "loss": 1.8124,
-      "step": 925000
-    },
-    {
-      "epoch": 13.32,
-      "learning_rate": 4.93574065708099e-06,
-      "loss": 1.811,
-      "step": 925500
-    },
-    {
-      "epoch": 13.33,
-      "learning_rate": 4.883813915328255e-06,
-      "loss": 1.8125,
-      "step": 926000
-    },
-    {
-      "epoch": 13.34,
-      "learning_rate": 4.83188717357552e-06,
-      "loss": 1.8155,
-      "step": 926500
-    },
-    {
-      "epoch": 13.34,
-      "learning_rate": 4.78006428530629e-06,
-      "loss": 1.8126,
-      "step": 927000
-    },
-    {
-      "epoch": 13.35,
-      "learning_rate": 4.728137543553555e-06,
-      "loss": 1.8134,
-      "step": 927500
-    },
-    {
-      "epoch": 13.36,
-      "learning_rate": 4.67621080180082e-06,
-      "loss": 1.8132,
-      "step": 928000
-    },
-    {
-      "epoch": 13.36,
-      "learning_rate": 4.6242840600480845e-06,
-      "loss": 1.8159,
-      "step": 928500
-    },
-    {
-      "epoch": 13.37,
-      "learning_rate": 4.572357318295349e-06,
-      "loss": 1.8134,
-      "step": 929000
-    },
-    {
-      "epoch": 13.38,
-      "learning_rate": 4.52053443002612e-06,
-      "loss": 1.8153,
-      "step": 929500
-    },
-    {
-      "epoch": 13.39,
-      "learning_rate": 4.468607688273385e-06,
-      "loss": 1.81,
-      "step": 930000
-    },
-    {
-      "epoch": 13.39,
-      "learning_rate": 4.416680946520649e-06,
-      "loss": 1.8081,
-      "step": 930500
-    },
-    {
-      "epoch": 13.4,
-      "learning_rate": 4.364754204767914e-06,
-      "loss": 1.8167,
-      "step": 931000
-    },
-    {
-      "epoch": 13.41,
-      "learning_rate": 4.312931316498684e-06,
-      "loss": 1.81,
-      "step": 931500
-    },
-    {
-      "epoch": 13.42,
-      "learning_rate": 4.261004574745948e-06,
-      "loss": 1.8106,
-      "step": 932000
-    },
-    {
-      "epoch": 13.42,
-      "learning_rate": 4.209077832993213e-06,
-      "loss": 1.812,
-      "step": 932500
-    },
-    {
-      "epoch": 13.43,
-      "learning_rate": 4.157151091240478e-06,
-      "loss": 1.8154,
-      "step": 933000
-    },
-    {
-      "epoch": 13.44,
-      "learning_rate": 4.105328202971248e-06,
-      "loss": 1.8144,
-      "step": 933500
-    },
-    {
-      "epoch": 13.44,
-      "learning_rate": 4.0534014612185135e-06,
-      "loss": 1.8145,
-      "step": 934000
-    },
-    {
-      "epoch": 13.45,
-      "learning_rate": 4.001474719465778e-06,
-      "loss": 1.8156,
-      "step": 934500
-    },
-    {
-      "epoch": 13.46,
-      "learning_rate": 3.949547977713042e-06,
-      "loss": 1.8142,
-      "step": 935000
-    },
-    {
-      "epoch": 13.47,
-      "learning_rate": 3.897725089443813e-06,
-      "loss": 1.8167,
-      "step": 935500
-    },
-    {
-      "epoch": 13.47,
-      "learning_rate": 3.845798347691078e-06,
-      "loss": 1.8122,
-      "step": 936000
-    },
-    {
-      "epoch": 13.48,
-      "learning_rate": 3.7938716059383423e-06,
-      "loss": 1.8148,
-      "step": 936500
-    },
-    {
-      "epoch": 13.49,
-      "learning_rate": 3.741944864185607e-06,
-      "loss": 1.8149,
-      "step": 937000
-    },
-    {
-      "epoch": 13.49,
-      "learning_rate": 3.6901219759163772e-06,
-      "loss": 1.8154,
-      "step": 937500
-    },
-    {
-      "epoch": 13.5,
-      "learning_rate": 3.6381952341636426e-06,
-      "loss": 1.8149,
-      "step": 938000
-    },
-    {
-      "epoch": 13.51,
-      "learning_rate": 3.5862684924109066e-06,
-      "loss": 1.8109,
-      "step": 938500
-    },
-    {
-      "epoch": 13.52,
-      "learning_rate": 3.534341750658172e-06,
-      "loss": 1.8105,
-      "step": 939000
-    },
-    {
-      "epoch": 13.52,
-      "learning_rate": 3.482415008905436e-06,
-      "loss": 1.8113,
-      "step": 939500
-    },
-    {
-      "epoch": 13.53,
-      "learning_rate": 3.4305921206362065e-06,
-      "loss": 1.812,
-      "step": 940000
-    },
-    {
-      "epoch": 13.54,
-      "learning_rate": 3.3786653788834714e-06,
-      "loss": 1.81,
-      "step": 940500
-    },
-    {
-      "epoch": 13.54,
-      "learning_rate": 3.3267386371307363e-06,
-      "loss": 1.8133,
-      "step": 941000
-    },
-    {
-      "epoch": 13.55,
-      "learning_rate": 3.2748118953780008e-06,
-      "loss": 1.8131,
-      "step": 941500
-    },
-    {
-      "epoch": 13.56,
-      "learning_rate": 3.2228851536252657e-06,
-      "loss": 1.8114,
-      "step": 942000
-    },
-    {
-      "epoch": 13.57,
-      "learning_rate": 3.171062265356036e-06,
-      "loss": 1.8116,
-      "step": 942500
-    },
-    {
-      "epoch": 13.57,
-      "learning_rate": 3.1191355236033006e-06,
-      "loss": 1.8144,
-      "step": 943000
-    },
-    {
-      "epoch": 13.58,
-      "learning_rate": 3.0672087818505655e-06,
-      "loss": 1.8101,
-      "step": 943500
-    },
-    {
-      "epoch": 13.59,
-      "learning_rate": 3.01528204009783e-06,
-      "loss": 1.8127,
-      "step": 944000
-    },
-    {
-      "epoch": 13.6,
-      "learning_rate": 2.9634591518286005e-06,
-      "loss": 1.8111,
-      "step": 944500
-    },
-    {
-      "epoch": 13.6,
-      "learning_rate": 2.911532410075865e-06,
-      "loss": 1.8082,
-      "step": 945000
-    },
-    {
-      "epoch": 13.61,
-      "learning_rate": 2.85960566832313e-06,
-      "loss": 1.8093,
-      "step": 945500
-    },
-    {
-      "epoch": 13.62,
-      "learning_rate": 2.8076789265703948e-06,
-      "loss": 1.8126,
-      "step": 946000
-    },
-    {
-      "epoch": 13.62,
-      "learning_rate": 2.755856038301165e-06,
-      "loss": 1.8086,
-      "step": 946500
-    },
-    {
-      "epoch": 13.63,
-      "learning_rate": 2.7039292965484297e-06,
-      "loss": 1.8113,
-      "step": 947000
-    },
-    {
-      "epoch": 13.64,
-      "learning_rate": 2.652002554795694e-06,
-      "loss": 1.8125,
-      "step": 947500
-    },
-    {
-      "epoch": 13.65,
-      "learning_rate": 2.600075813042959e-06,
-      "loss": 1.8107,
-      "step": 948000
-    },
-    {
-      "epoch": 13.65,
-      "learning_rate": 2.5481490712902236e-06,
-      "loss": 1.8119,
-      "step": 948500
-    },
-    {
-      "epoch": 13.66,
-      "learning_rate": 2.496326183020994e-06,
-      "loss": 1.8121,
-      "step": 949000
-    },
-    {
-      "epoch": 13.67,
-      "learning_rate": 2.4443994412682585e-06,
-      "loss": 1.8153,
-      "step": 949500
-    },
-    {
-      "epoch": 13.67,
-      "learning_rate": 2.392472699515524e-06,
-      "loss": 1.811,
-      "step": 950000
-    },
-    {
-      "epoch": 13.68,
-      "learning_rate": 2.3405459577627883e-06,
-      "loss": 1.814,
-      "step": 950500
-    },
-    {
-      "epoch": 13.69,
-      "learning_rate": 2.288723069493559e-06,
-      "loss": 1.8147,
-      "step": 951000
-    },
-    {
-      "epoch": 13.7,
-      "learning_rate": 2.2367963277408233e-06,
-      "loss": 1.8152,
-      "step": 951500
-    },
-    {
-      "epoch": 13.7,
-      "learning_rate": 2.184869585988088e-06,
-      "loss": 1.8114,
-      "step": 952000
-    },
-    {
-      "epoch": 13.71,
-      "learning_rate": 2.1329428442353527e-06,
-      "loss": 1.8092,
-      "step": 952500
-    },
-    {
-      "epoch": 13.72,
-      "learning_rate": 2.0810161024826176e-06,
-      "loss": 1.809,
-      "step": 953000
-    },
-    {
-      "epoch": 13.72,
-      "learning_rate": 2.0290893607298825e-06,
-      "loss": 1.8162,
-      "step": 953500
-    },
-    {
-      "epoch": 13.73,
-      "learning_rate": 1.9772664724606526e-06,
-      "loss": 1.8088,
-      "step": 954000
-    },
-    {
-      "epoch": 13.74,
-      "learning_rate": 1.9253397307079175e-06,
-      "loss": 1.8148,
-      "step": 954500
-    },
-    {
-      "epoch": 13.75,
-      "learning_rate": 1.8734129889551821e-06,
-      "loss": 1.8101,
-      "step": 955000
-    },
-    {
-      "epoch": 13.75,
-      "learning_rate": 1.8214862472024468e-06,
-      "loss": 1.813,
-      "step": 955500
-    },
-    {
-      "epoch": 13.76,
-      "learning_rate": 1.769663358933217e-06,
-      "loss": 1.8143,
-      "step": 956000
-    },
-    {
-      "epoch": 13.77,
-      "learning_rate": 1.7177366171804818e-06,
-      "loss": 1.8085,
-      "step": 956500
-    },
-    {
-      "epoch": 13.78,
-      "learning_rate": 1.6658098754277467e-06,
-      "loss": 1.8148,
-      "step": 957000
-    },
-    {
-      "epoch": 13.78,
-      "learning_rate": 1.6138831336750114e-06,
-      "loss": 1.8081,
-      "step": 957500
-    },
-    {
-      "epoch": 13.79,
-      "learning_rate": 1.561956391922276e-06,
-      "loss": 1.8114,
-      "step": 958000
-    },
-    {
-      "epoch": 13.8,
-      "learning_rate": 1.5101335036530463e-06,
-      "loss": 1.8083,
-      "step": 958500
-    },
-    {
-      "epoch": 13.8,
-      "learning_rate": 1.458206761900311e-06,
-      "loss": 1.8137,
-      "step": 959000
-    },
-    {
-      "epoch": 13.81,
-      "learning_rate": 1.406280020147576e-06,
-      "loss": 1.815,
-      "step": 959500
-    },
-    {
-      "epoch": 13.82,
-      "learning_rate": 1.3543532783948406e-06,
-      "loss": 1.813,
-      "step": 960000
-    },
-    {
-      "epoch": 13.83,
-      "learning_rate": 1.302530390125611e-06,
-      "loss": 1.812,
-      "step": 960500
-    },
-    {
-      "epoch": 13.83,
-      "learning_rate": 1.2506036483728756e-06,
-      "loss": 1.8105,
-      "step": 961000
-    },
-    {
-      "epoch": 13.84,
-      "learning_rate": 1.1986769066201403e-06,
-      "loss": 1.8142,
-      "step": 961500
-    },
-    {
-      "epoch": 13.85,
-      "learning_rate": 1.146750164867405e-06,
-      "loss": 1.8101,
-      "step": 962000
-    },
-    {
-      "epoch": 13.85,
-      "learning_rate": 1.0948234231146699e-06,
-      "loss": 1.8141,
-      "step": 962500
-    },
-    {
-      "epoch": 13.86,
-      "learning_rate": 1.0430005348454401e-06,
-      "loss": 1.8125,
-      "step": 963000
-    },
-    {
-      "epoch": 13.87,
-      "learning_rate": 9.910737930927048e-07,
-      "loss": 1.8128,
-      "step": 963500
-    },
-    {
-      "epoch": 13.88,
-      "learning_rate": 9.391470513399695e-07,
-      "loss": 1.8126,
-      "step": 964000
-    },
-    {
-      "epoch": 13.88,
-      "learning_rate": 8.872203095872344e-07,
-      "loss": 1.8131,
-      "step": 964500
-    },
-    {
-      "epoch": 13.89,
-      "learning_rate": 8.353974213180045e-07,
-      "loss": 1.8105,
-      "step": 965000
-    },
-    {
-      "epoch": 13.9,
-      "learning_rate": 7.834706795652694e-07,
-      "loss": 1.8119,
-      "step": 965500
-    },
-    {
-      "epoch": 13.9,
-      "learning_rate": 7.315439378125341e-07,
-      "loss": 1.8076,
-      "step": 966000
-    },
-    {
-      "epoch": 13.91,
-      "learning_rate": 6.796171960597989e-07,
-      "loss": 1.809,
-      "step": 966500
-    },
-    {
-      "epoch": 13.92,
-      "learning_rate": 6.277943077905691e-07,
-      "loss": 1.8124,
-      "step": 967000
-    },
-    {
-      "epoch": 13.93,
-      "learning_rate": 5.758675660378338e-07,
-      "loss": 1.811,
-      "step": 967500
-    },
-    {
-      "epoch": 13.93,
-      "learning_rate": 5.239408242850986e-07,
-      "loss": 1.8105,
-      "step": 968000
-    },
-    {
-      "epoch": 13.94,
-      "learning_rate": 4.7201408253236337e-07,
-      "loss": 1.8122,
-      "step": 968500
-    },
-    {
-      "epoch": 13.95,
-      "learning_rate": 4.200873407796281e-07,
-      "loss": 1.8113,
-      "step": 969000
-    },
-    {
-      "epoch": 13.96,
-      "learning_rate": 3.682644525103984e-07,
-      "loss": 1.8112,
-      "step": 969500
-    },
-    {
-      "epoch": 13.96,
-      "learning_rate": 3.1633771075766313e-07,
-      "loss": 1.8096,
-      "step": 970000
-    },
-    {
-      "epoch": 13.97,
-      "learning_rate": 2.6441096900492787e-07,
-      "loss": 1.8133,
-      "step": 970500
-    },
-    {
-      "epoch": 13.98,
-      "learning_rate": 2.1248422725219261e-07,
-      "loss": 1.8126,
-      "step": 971000
-    },
-    {
-      "epoch": 13.98,
-      "learning_rate": 1.6066133898296286e-07,
-      "loss": 1.8129,
-      "step": 971500
-    },
-    {
-      "epoch": 13.99,
-      "learning_rate": 1.087345972302276e-07,
-      "loss": 1.8086,
-      "step": 972000
-    },
-    {
-      "epoch": 14.0,
-      "learning_rate": 5.6807855477492356e-08,
-      "loss": 1.8122,
-      "step": 972500
-    },
-    {
-      "epoch": 14.0,
-      "eval_accuracy": 0.6519198053032371,
-      "eval_loss": 1.6776695251464844,
-      "eval_runtime": 646.1912,
-      "eval_samples_per_second": 834.021,
-      "eval_steps_per_second": 34.751,
-      "step": 972622
-    }
-  ],
-  "max_steps": 972622,
-  "num_train_epochs": 14,
-  "total_flos": 6.804216664343708e+18,
-  "trial_name": null,
-  "trial_params": null
-}